Daily Papers

Masked Language Models (MLMs) have proven to be effective for second-pass rescoring in Automatic Speech Recognition (ASR) systems. In this work, we propose Masked Audio Text Encoder (MATE), a multi-modal masked language model rescorer which incorporates acoustic representations into the input space of MLM. We adopt contrastive learning for effectively aligning the modalities by learning shared representations. We show that using a multi-modal rescorer is beneficial for domain generalization of the ASR system when target domain data is unavailable. MATE reduces word error rate (WER) by 4%-16% on in-domain, and 3%-7% on out-of-domain datasets, over the text-only baseline. Additionally, with very limited amount of training data (0.8 hours), MATE achieves a WER reduction of 8%-23% over the first-pass baseline.

We propose EnCLAP, a novel framework for automated audio captioning. EnCLAP employs two acoustic representation models, EnCodec and CLAP, along with a pretrained language model, BART. We also introduce a new training objective called masked codec modeling that improves acoustic awareness of the pretrained language model. Experimental results on AudioCaps and Clotho demonstrate that our model surpasses the performance of baseline models. Source code will be available at https://github.com/jaeyeonkim99/EnCLAP . An online demo is available at https://huggingface.co/spaces/enclap-team/enclap .

The goal of universal audio representation learning is to obtain foundational models that can be used for a variety of downstream tasks involving speech, music or environmental sounds. To approach this problem, methods inspired by self-supervised models from NLP, like BERT, are often used and adapted to audio. These models rely on the discrete nature of text, hence adopting this type of approach for audio processing requires either a change in the learning objective or mapping the audio signal to a set of discrete classes. In this work, we explore the use of EnCodec, a neural audio codec, to generate discrete targets for learning an universal audio model based on a masked autoencoder (MAE). We evaluate this approach, which we call EncodecMAE, on a wide range of audio tasks spanning speech, music and environmental sounds, achieving performances comparable or better than leading audio representation models.

This paper studies a simple extension of image-based Masked Autoencoders (MAE) to self-supervised representation learning from audio spectrograms. Following the Transformer encoder-decoder design in MAE, our Audio-MAE first encodes audio spectrogram patches with a high masking ratio, feeding only the non-masked tokens through encoder layers. The decoder then re-orders and decodes the encoded context padded with mask tokens, in order to reconstruct the input spectrogram. We find it beneficial to incorporate local window attention in the decoder, as audio spectrograms are highly correlated in local time and frequency bands. We then fine-tune the encoder with a lower masking ratio on target datasets. Empirically, Audio-MAE sets new state-of-the-art performance on six audio and speech classification tasks, outperforming other recent models that use external supervised pre-training. The code and models will be at https://github.com/facebookresearch/AudioMAE.

Automated Audio Captioning (AAC) involves generating natural language descriptions of audio content, using encoder-decoder architectures. An audio encoder produces audio embeddings fed to a decoder, usually a Transformer decoder, for caption generation. In this work, we describe our model, which novelty, compared to existing models, lies in the use of a ConvNeXt architecture as audio encoder, adapted from the vision domain to audio classification. This model, called CNext-trans, achieved state-of-the-art scores on the AudioCaps (AC) dataset and performed competitively on Clotho (CL), while using four to forty times fewer parameters than existing models. We examine potential biases in the AC dataset due to its origin from AudioSet by investigating unbiased encoder's impact on performance. Using the well-known PANN's CNN14, for instance, as an unbiased encoder, we observed a 1.7% absolute reduction in SPIDEr score (where higher scores indicate better performance). To improve cross-dataset performance, we conducted experiments by combining multiple AAC datasets (AC, CL, MACS, WavCaps) for training. Although this strategy enhanced overall model performance across datasets, it still fell short compared to models trained specifically on a single target dataset, indicating the absence of a one-size-fits-all model. To mitigate performance gaps between datasets, we introduced a Task Embedding (TE) token, allowing the model to identify the source dataset for each input sample. We provide insights into the impact of these TEs on both the form (words) and content (sound event types) of the generated captions. The resulting model, named CoNeTTE, an unbiased CNext-trans model enriched with dataset-specific Task Embeddings, achieved SPIDEr scores of 44.1% and 30.5% on AC and CL, respectively. Code available: https://github.com/Labbeti/conette-audio-captioning.

Video-to-audio (V2A) generation leverages visual-only video features to render plausible sounds that match the scene. Importantly, the generated sound onsets should match the visual actions that are aligned with them, otherwise unnatural synchronization artifacts arise. Recent works have explored the progression of conditioning sound generators on still images and then video features, focusing on quality and semantic matching while ignoring synchronization, or by sacrificing some amount of quality to focus on improving synchronization only. In this work, we propose a V2A generative model, named MaskVAT, that interconnects a full-band high-quality general audio codec with a sequence-to-sequence masked generative model. This combination allows modeling both high audio quality, semantic matching, and temporal synchronicity at the same time. Our results show that, by combining a high-quality codec with the proper pre-trained audio-visual features and a sequence-to-sequence parallel structure, we are able to yield highly synchronized results on one hand, whilst being competitive with the state of the art of non-codec generative audio models. Sample videos and generated audios are available at https://maskvat.github.io .

We present the first approach to automated audio captioning. We employ an encoder-decoder scheme with an alignment model in between. The input to the encoder is a sequence of log mel-band energies calculated from an audio file, while the output is a sequence of words, i.e. a caption. The encoder is a multi-layered, bi-directional gated recurrent unit (GRU) and the decoder a multi-layered GRU with a classification layer connected to the last GRU of the decoder. The classification layer and the alignment model are fully connected layers with shared weights between timesteps. The proposed method is evaluated using data drawn from a commercial sound effects library, ProSound Effects. The resulting captions were rated through metrics utilized in machine translation and image captioning fields. Results from metrics show that the proposed method can predict words appearing in the original caption, but not always correctly ordered.

In the rapidly evolving field of speech generative models, there is a pressing need to ensure audio authenticity against the risks of voice cloning. We present AudioSeal, the first audio watermarking technique designed specifically for localized detection of AI-generated speech. AudioSeal employs a generator/detector architecture trained jointly with a localization loss to enable localized watermark detection up to the sample level, and a novel perceptual loss inspired by auditory masking, that enables AudioSeal to achieve better imperceptibility. AudioSeal achieves state-of-the-art performance in terms of robustness to real life audio manipulations and imperceptibility based on automatic and human evaluation metrics. Additionally, AudioSeal is designed with a fast, single-pass detector, that significantly surpasses existing models in speed - achieving detection up to two orders of magnitude faster, making it ideal for large-scale and real-time applications.

Recent advances in generative models that iteratively synthesize audio clips sparked great success to text-to-audio synthesis (TTA), but with the cost of slow synthesis speed and heavy computation. Although there have been attempts to accelerate the iterative procedure, high-quality TTA systems remain inefficient due to hundreds of iterations required in the inference phase and large amount of model parameters. To address the challenges, we propose SpecMaskGIT, a light-weighted, efficient yet effective TTA model based on the masked generative modeling of spectrograms. First, SpecMaskGIT synthesizes a realistic 10s audio clip by less than 16 iterations, an order-of-magnitude less than previous iterative TTA methods.As a discrete model, SpecMaskGIT outperforms larger VQ-Diffusion and auto-regressive models in the TTA benchmark, while being real-time with only 4 CPU cores or even 30x faster with a GPU. Next, built upon a latent space of Mel-spectrogram, SpecMaskGIT has a wider range of applications (e.g., the zero-shot bandwidth extension) than similar methods built on the latent wave domain. Moreover, we interpret SpecMaskGIT as a generative extension to previous discriminative audio masked Transformers, and shed light on its audio representation learning potential. We hope our work inspires the exploration of masked audio modeling toward further diverse scenarios.

In this paper, we first extend the recent Masked Auto-Encoder (MAE) model from a single modality to audio-visual multi-modalities. Subsequently, we propose the Contrastive Audio-Visual Masked Auto-Encoder (CAV-MAE) by combining contrastive learning and masked data modeling, two major self-supervised learning frameworks, to learn a joint and coordinated audio-visual representation. Our experiments show that the contrastive audio-visual correspondence learning objective not only enables the model to perform audio-visual retrieval tasks, but also helps the model learn a better joint representation. As a result, our fully self-supervised pretrained CAV-MAE achieves a new SOTA accuracy of 65.9% on VGGSound, and is comparable with the previous best supervised pretrained model on AudioSet in the audio-visual event classification task. Code and pretrained models are at https://github.com/yuangongnd/cav-mae.

Masked speech modeling (MSM) methods such as wav2vec2 or w2v-BERT learn representations over speech frames which are randomly masked within an utterance. While these methods improve performance of Automatic Speech Recognition (ASR) systems, they have one major limitation. They treat all unsupervised speech samples with equal weight, which hinders learning as not all samples have relevant information to learn meaningful representations. In this work, we address this limitation. We propose ask2mask (ATM), a novel approach to focus on specific samples during MSM pre-training. ATM employs an external ASR model or scorer to weight unsupervised input samples in two different ways: 1) A fine-grained data selection is performed by masking over the highly confident input frames as chosen by the scorer. This allows the model to learn meaningful representations. 2) ATM is further extended to focus at utterance-level by weighting the final MSM loss with the utterance-level confidence score. We conduct fine-tuning experiments on two well-benchmarked corpora: LibriSpeech (matching the pre-training data) and Commonvoice, TED-LIUM, AMI and CHiME-6 (not matching the pre-training data). The results substantiate the efficacy of ATM on significantly improving the recognition performance under mismatched conditions (up to 11.6\% relative over published results and upto 4.46\% relative over our internal baseline) while still yielding modest improvements under matched conditions.

Large language models (LLMs) have significantly advanced audio processing through audio codecs that convert audio into discrete tokens, enabling the application of language modelling techniques to audio data. However, traditional codecs often operate at high bitrates or within narrow domains such as speech and lack the semantic clues required for efficient language modelling. Addressing these challenges, we introduce SemantiCodec, a novel codec designed to compress audio into fewer than a hundred tokens per second across diverse audio types, including speech, general audio, and music, without compromising quality. SemantiCodec features a dual-encoder architecture: a semantic encoder using a self-supervised AudioMAE, discretized using k-means clustering on extensive audio data, and an acoustic encoder to capture the remaining details. The semantic and acoustic encoder outputs are used to reconstruct audio via a diffusion-model-based decoder. SemantiCodec is presented in three variants with token rates of 25, 50, and 100 per second, supporting a range of ultra-low bit rates between 0.31 kbps and 1.43 kbps. Experimental results demonstrate that SemantiCodec significantly outperforms the state-of-the-art Descript codec on reconstruction quality. Our results also suggest that SemantiCodec contains significantly richer semantic information than all evaluated audio codecs, even at significantly lower bitrates. Our code and demos are available at https://haoheliu.github.io/SemantiCodec/.

The recent large-scale text-to-speech (TTS) systems are usually grouped as autoregressive and non-autoregressive systems. The autoregressive systems implicitly model duration but exhibit certain deficiencies in robustness and lack of duration controllability. Non-autoregressive systems require explicit alignment information between text and speech during training and predict durations for linguistic units (e.g. phone), which may compromise their naturalness. In this paper, we introduce Masked Generative Codec Transformer (MaskGCT), a fully non-autoregressive TTS model that eliminates the need for explicit alignment information between text and speech supervision, as well as phone-level duration prediction. MaskGCT is a two-stage model: in the first stage, the model uses text to predict semantic tokens extracted from a speech self-supervised learning (SSL) model, and in the second stage, the model predicts acoustic tokens conditioned on these semantic tokens. MaskGCT follows the mask-and-predict learning paradigm. During training, MaskGCT learns to predict masked semantic or acoustic tokens based on given conditions and prompts. During inference, the model generates tokens of a specified length in a parallel manner. Experiments with 100K hours of in-the-wild speech demonstrate that MaskGCT outperforms the current state-of-the-art zero-shot TTS systems in terms of quality, similarity, and intelligibility. Audio samples are available at https://maskgct.github.io/.

We propose EMAGE, a framework to generate full-body human gestures from audio and masked gestures, encompassing facial, local body, hands, and global movements. To achieve this, we first introduce BEAT2 (BEAT-SMPLX-FLAME), a new mesh-level holistic co-speech dataset. BEAT2 combines MoShed SMPLX body with FLAME head parameters and further refines the modeling of head, neck, and finger movements, offering a community-standardized, high-quality 3D motion captured dataset. EMAGE leverages masked body gesture priors during training to boost inference performance. It involves a Masked Audio Gesture Transformer, facilitating joint training on audio-to-gesture generation and masked gesture reconstruction to effectively encode audio and body gesture hints. Encoded body hints from masked gestures are then separately employed to generate facial and body movements. Moreover, EMAGE adaptively merges speech features from the audio's rhythm and content and utilizes four compositional VQ-VAEs to enhance the results' fidelity and diversity. Experiments demonstrate that EMAGE generates holistic gestures with state-of-the-art performance and is flexible in accepting predefined spatial-temporal gesture inputs, generating complete, audio-synchronized results. Our code and dataset are available at https://pantomatrix.github.io/EMAGE/

We propose WHISPER-GPT: A generative large language model (LLM) for speech and music that allows us to work with continuous audio representations and discrete tokens simultaneously as part of a single architecture. There has been a huge surge in generative audio, speech, and music models that utilize discrete audio tokens derived from neural compression algorithms, e.g. ENCODEC. However, one of the major drawbacks of this approach is handling the context length. It blows up for high-fidelity generative architecture if one has to account for all the audio contents at various frequencies for the next token prediction. By combining continuous audio representation like the spectrogram and discrete acoustic tokens, we retain the best of both worlds: Have all the information needed from the audio at a specific time instance in a single token, yet allow LLM to predict the future token to allow for sampling and other benefits discrete space provides. We show how our architecture improves the perplexity and negative log-likelihood scores for the next token prediction compared to a token-based LLM for speech and music.

In speech technologies, speaker's voice representation is used in many applications such as speech recognition, voice conversion, speech synthesis and, obviously, user authentication. Modern vocal representations of the speaker are based on neural embeddings. In addition to the targeted information, these representations usually contain sensitive information about the speaker, like the age, sex, physical state, education level or ethnicity. In order to allow the user to choose which information to protect, we introduce in this paper the concept of attribute-driven privacy preservation in speaker voice representation. It allows a person to hide one or more personal aspects to a potential malicious interceptor and to the application provider. As a first solution to this concept, we propose to use an adversarial autoencoding method that disentangles in the voice representation a given speaker attribute thus allowing its concealment. We focus here on the sex attribute for an Automatic Speaker Verification (ASV) task. Experiments carried out using the VoxCeleb datasets have shown that the proposed method enables the concealment of this attribute while preserving ASV ability.

Automated Audio Captioning (AAC) is the task of generating natural language descriptions given an audio stream. A typical AAC system requires manually curated training data of audio segments and corresponding text caption annotations. The creation of these audio-caption pairs is costly, resulting in general data scarcity for the task. In this work, we address this major limitation and propose an approach to train AAC systems using only text. Our approach leverages the multimodal space of contrastively trained audio-text models, such as CLAP. During training, a decoder generates captions conditioned on the pretrained CLAP text encoder. During inference, the text encoder is replaced with the pretrained CLAP audio encoder. To bridge the modality gap between text and audio embeddings, we propose the use of noise injection or a learnable adapter, during training. We find that the proposed text-only framework performs competitively with state-of-the-art models trained with paired audio, showing that efficient text-to-audio transfer is possible. Finally, we showcase both stylized audio captioning and caption enrichment while training without audio or human-created text captions.

In the realm of audio watermarking, it is challenging to simultaneously encode imperceptible messages while enhancing the message capacity and robustness. Although recent advancements in deep learning-based methods bolster the message capacity and robustness over traditional methods, the encoded messages introduce audible artefacts that restricts their usage in professional settings. In this study, we introduce three key innovations. Firstly, our work is the first deep learning-based model to integrate psychoacoustic model based thresholding to achieve imperceptible watermarks. Secondly, we introduce psuedo-differentiable compression layers, enhancing the robustness of our watermarking algorithm. Lastly, we introduce a method to eliminate the need for perceptual losses, enabling us to achieve SOTA in both robustness as well as imperceptible watermarking. Our contributions lead us to SilentCipher, a model enabling users to encode messages within audio signals sampled at 44.1kHz.

Audio self-supervised learning (SSL) pre-training, which aims to learn good representations from unlabeled audio, has made remarkable progress. However, the extensive computational demands during pre-training pose a significant barrier to the potential application and optimization of audio SSL models. In this paper, inspired by the success of data2vec 2.0 in image modality and Audio-MAE in audio modality, we introduce Efficient Audio Transformer (EAT) to further improve the effectiveness and efficiency in audio SSL. The proposed EAT adopts the bootstrap self-supervised training paradigm to the audio domain. A novel Utterance-Frame Objective (UFO) is designed to enhance the modeling capability of acoustic events. Furthermore, we reveal that the masking strategy is critical in audio SSL pre-training, and superior audio representations can be obtained with large inverse block masks. Experiment results demonstrate that EAT achieves state-of-the-art (SOTA) performance on a range of audio-related tasks, including AudioSet (AS-2M, AS-20K), ESC-50, and SPC-2, along with a significant pre-training speedup up to ~15x compared to existing audio SSL models.

This technical report describes the MERaLiON Speech Encoder, a foundation model designed to support a wide range of downstream speech applications. Developed as part of Singapore's National Multimodal Large Language Model Programme, the MERaLiON Speech Encoder is tailored to address the speech processing needs in Singapore and the surrounding Southeast Asian region. The model currently supports mainly English, including the variety spoken in Singapore. We are actively expanding our datasets to gradually cover other languages in subsequent releases. The MERaLiON Speech Encoder was pre-trained from scratch on 200K hours of unlabelled speech data using a self-supervised learning approach based on masked language modelling. We describe our training procedure and hyperparameter tuning experiments in detail below. Our evaluation demonstrates improvements to spontaneous and Singapore speech benchmarks for speech recognition, while remaining competitive to other state-of-the-art speech encoders across ten other speech tasks. We commit to releasing our model, supporting broader research endeavours, both in Singapore and beyond.

This paper presents VoiceLDM, a model designed to produce audio that accurately follows two distinct natural language text prompts: the description prompt and the content prompt. The former provides information about the overall environmental context of the audio, while the latter conveys the linguistic content. To achieve this, we adopt a text-to-audio (TTA) model based on latent diffusion models and extend its functionality to incorporate an additional content prompt as a conditional input. By utilizing pretrained contrastive language-audio pretraining (CLAP) and Whisper, VoiceLDM is trained on large amounts of real-world audio without manual annotations or transcriptions. Additionally, we employ dual classifier-free guidance to further enhance the controllability of VoiceLDM. Experimental results demonstrate that VoiceLDM is capable of generating plausible audio that aligns well with both input conditions, even surpassing the speech intelligibility of the ground truth audio on the AudioCaps test set. Furthermore, we explore the text-to-speech (TTS) and zero-shot text-to-audio capabilities of VoiceLDM and show that it achieves competitive results. Demos and code are available at https://voiceldm.github.io.

In the domain of audio processing, Transfer Learning has facilitated the rise of Self-Supervised Learning and Zero-Shot Learning techniques. These approaches have led to the development of versatile models capable of tackling a wide array of tasks, while delivering state-of-the-art performance. However, current models inherently lack the capacity to produce the requisite language for open-ended tasks, such as Audio Captioning or Audio Question & Answering. We introduce Pengi, a novel Audio Language Model that leverages Transfer Learning by framing all audio tasks as text-generation tasks. It takes as input, an audio recording, and text, and generates free-form text as output. The input audio is represented as a sequence of continuous embeddings by an audio encoder. A text encoder does the same for the corresponding text input. Both sequences are combined as a prefix to prompt a pre-trained frozen language model. The unified architecture of Pengi enables open-ended tasks and close-ended tasks without any additional fine-tuning or task-specific extensions. When evaluated on 22 downstream tasks, our approach yields state-of-the-art performance in several of them. Our results show that connecting language models with audio models is a major step towards general-purpose audio understanding

In this work, we propose a Multi-Window Masked Autoencoder (MW-MAE) fitted with a novel Multi-Window Multi-Head Attention (MW-MHA) module that facilitates the modelling of local-global interactions in every decoder transformer block through attention heads of several distinct local and global windows. Empirical results on ten downstream audio tasks show that MW-MAEs consistently outperform standard MAEs in overall performance and learn better general-purpose audio representations, along with demonstrating considerably better scaling characteristics. Investigating attention distances and entropies reveals that MW-MAE encoders learn heads with broader local and global attention. Analyzing attention head feature representations through Projection Weighted Canonical Correlation Analysis (PWCCA) shows that attention heads with the same window sizes across the decoder layers of the MW-MAE learn correlated feature representations which enables each block to independently capture local and global information, leading to a decoupled decoder feature hierarchy. Code for feature extraction and downstream experiments along with pre-trained models will be released publically.

BERT (Bidirectional Encoder Representations from Transformers) has revolutionized the field of natural language processing through its exceptional performance on numerous tasks. Yet, the majority of researchers have mainly concentrated on enhancements related to the model structure, such as relative position embedding and more efficient attention mechanisms. Others have delved into pretraining tricks associated with Masked Language Modeling, including whole word masking. DeBERTa introduced an enhanced decoder adapted for BERT's encoder model for pretraining, proving to be highly effective. We argue that the design and research around enhanced masked language modeling decoders have been underappreciated. In this paper, we propose several designs of enhanced decoders and introduce DrBERT (Decoder-refined BERT), a novel method for modeling training. Typically, a pretrained BERT model is fine-tuned for specific Natural Language Understanding (NLU) tasks. In our approach, we utilize the original BERT model as the encoder, making only changes to the decoder without altering the encoder. This approach does not necessitate extensive modifications to the model's architecture and can be seamlessly integrated into existing fine-tuning pipelines and services, offering an efficient and effective enhancement strategy. Compared to other methods, while we also incur a moderate training cost for the decoder during the pretraining process, our approach does not introduce additional training costs during the fine-tuning phase. We test multiple enhanced decoder structures after pretraining and evaluate their performance on the GLUE benchmark. Our results demonstrate that DrBERT, having only undergone subtle refinements to the model structure during pretraining, significantly enhances model performance without escalating the inference time and serving budget.

We introduce a state-of-the-art real-time, high-fidelity, audio codec leveraging neural networks. It consists in a streaming encoder-decoder architecture with quantized latent space trained in an end-to-end fashion. We simplify and speed-up the training by using a single multiscale spectrogram adversary that efficiently reduces artifacts and produce high-quality samples. We introduce a novel loss balancer mechanism to stabilize training: the weight of a loss now defines the fraction of the overall gradient it should represent, thus decoupling the choice of this hyper-parameter from the typical scale of the loss. Finally, we study how lightweight Transformer models can be used to further compress the obtained representation by up to 40%, while staying faster than real time. We provide a detailed description of the key design choices of the proposed model including: training objective, architectural changes and a study of various perceptual loss functions. We present an extensive subjective evaluation (MUSHRA tests) together with an ablation study for a range of bandwidths and audio domains, including speech, noisy-reverberant speech, and music. Our approach is superior to the baselines methods across all evaluated settings, considering both 24 kHz monophonic and 48 kHz stereophonic audio. Code and models are available at github.com/facebookresearch/encodec.

Visual Speech Recognition (VSR) differs from the common perception tasks as it requires deeper reasoning over the video sequence, even by human experts. Despite the recent advances in VSR, current approaches rely on labeled data to fully train or finetune their models predicting the target speech. This hinders their ability to generalize well beyond the training set and leads to performance degeneration under out-of-distribution challenging scenarios. Unlike previous works that involve auxiliary losses or complex training procedures and architectures, we propose a simple approach, named Lip2Vec that is based on learning a prior model. Given a robust visual speech encoder, this network maps the encoded latent representations of the lip sequence to their corresponding latents from the audio pair, which are sufficiently invariant for effective text decoding. The generated audio representation is then decoded to text using an off-the-shelf Audio Speech Recognition (ASR) model. The proposed model compares favorably with fully-supervised learning methods on the LRS3 dataset achieving 26 WER. Unlike SoTA approaches, our model keeps a reasonable performance on the VoxCeleb test set. We believe that reprogramming the VSR as an ASR task narrows the performance gap between the two and paves the way for more flexible formulations of lip reading.

Audio and music generation have emerged as crucial tasks in many applications, yet existing approaches face significant limitations: they operate in isolation without unified capabilities across modalities, suffer from scarce high-quality, multi-modal training data, and struggle to effectively integrate diverse inputs. In this work, we propose AudioX, a unified Diffusion Transformer model for Anything-to-Audio and Music Generation. Unlike previous domain-specific models, AudioX can generate both general audio and music with high quality, while offering flexible natural language control and seamless processing of various modalities including text, video, image, music, and audio. Its key innovation is a multi-modal masked training strategy that masks inputs across modalities and forces the model to learn from masked inputs, yielding robust and unified cross-modal representations. To address data scarcity, we curate two comprehensive datasets: vggsound-caps with 190K audio captions based on the VGGSound dataset, and V2M-caps with 6 million music captions derived from the V2M dataset. Extensive experiments demonstrate that AudioX not only matches or outperforms state-of-the-art specialized models, but also offers remarkable versatility in handling diverse input modalities and generation tasks within a unified architecture. The code and datasets will be available at https://zeyuet.github.io/AudioX/

Recent advancements in large language models (LLMs) have revolutionized various domains, bringing significant progress and new opportunities. Despite progress in speech-related tasks, LLMs have not been sufficiently explored in multi-talker scenarios. In this work, we present a pioneering effort to investigate the capability of LLMs in transcribing speech in multi-talker environments, following versatile instructions related to multi-talker automatic speech recognition (ASR), target talker ASR, and ASR based on specific talker attributes such as sex, occurrence order, language, and keyword spoken. Our approach utilizes WavLM and Whisper encoder to extract multi-faceted speech representations that are sensitive to speaker characteristics and semantic context. These representations are then fed into an LLM fine-tuned using LoRA, enabling the capabilities for speech comprehension and transcription. Comprehensive experiments reveal the promising performance of our proposed system, MT-LLM, in cocktail party scenarios, highlighting the potential of LLM to handle speech-related tasks based on user instructions in such complex settings.

We propose Fast Language-Audio Pre-training (FLAP), a self-supervised approach that efficiently and effectively learns aligned audio and language representations through masking, contrastive learning and reconstruction. For efficiency, FLAP randomly drops audio spectrogram tokens, focusing solely on the remaining ones for self-supervision. Through inter-modal contrastive learning, FLAP learns to align paired audio and text representations in a shared latent space. Notably, FLAP leverages multiple augmented views via masking for inter-modal contrast and learns to reconstruct the masked portion of audio tokens. Moreover, FLAP leverages large language models (LLMs) to augment the text inputs, contributing to improved performance. These approaches lead to more robust and informative audio-text representations, enabling FLAP to achieve state-of-the-art (SoTA) performance on audio-text retrieval tasks on AudioCaps (achieving 53.0% R@1) and Clotho (achieving 25.5% R@1).

We introduce MAGNeT, a masked generative sequence modeling method that operates directly over several streams of audio tokens. Unlike prior work, MAGNeT is comprised of a single-stage, non-autoregressive transformer. During training, we predict spans of masked tokens obtained from a masking scheduler, while during inference we gradually construct the output sequence using several decoding steps. To further enhance the quality of the generated audio, we introduce a novel rescoring method in which, we leverage an external pre-trained model to rescore and rank predictions from MAGNeT, which will be then used for later decoding steps. Lastly, we explore a hybrid version of MAGNeT, in which we fuse between autoregressive and non-autoregressive models to generate the first few seconds in an autoregressive manner while the rest of the sequence is being decoded in parallel. We demonstrate the efficiency of MAGNeT for the task of text-to-music and text-to-audio generation and conduct an extensive empirical evaluation, considering both objective metrics and human studies. The proposed approach is comparable to the evaluated baselines, while being significantly faster (x7 faster than the autoregressive baseline). Through ablation studies and analysis, we shed light on the importance of each of the components comprising MAGNeT, together with pointing to the trade-offs between autoregressive and non-autoregressive modeling, considering latency, throughput, and generation quality. Samples are available on our demo page https://pages.cs.huji.ac.il/adiyoss-lab/MAGNeT.

Over the past several years, the synchronization between audio and visual signals has been leveraged to learn richer audio-visual representations. Aided by the large availability of unlabeled videos, many unsupervised training frameworks have demonstrated impressive results in various downstream audio and video tasks. Recently, Masked Audio-Video Learners (MAViL) has emerged as a state-of-the-art audio-video pre-training framework. MAViL couples contrastive learning with masked autoencoding to jointly reconstruct audio spectrograms and video frames by fusing information from both modalities. In this paper, we study the potential synergy between diffusion models and MAViL, seeking to derive mutual benefits from these two frameworks. The incorporation of diffusion into MAViL, combined with various training efficiency methodologies that include the utilization of a masking ratio curriculum and adaptive batch sizing, results in a notable 32% reduction in pre-training Floating-Point Operations (FLOPS) and an 18% decrease in pre-training wall clock time. Crucially, this enhanced efficiency does not compromise the model's performance in downstream audio-classification tasks when compared to MAViL's performance.

Large diffusion models have been successful in text-to-audio (T2A) synthesis tasks, but they often suffer from common issues such as semantic misalignment and poor temporal consistency due to limited natural language understanding and data scarcity. Additionally, 2D spatial structures widely used in T2A works lead to unsatisfactory audio quality when generating variable-length audio samples since they do not adequately prioritize temporal information. To address these challenges, we propose Make-an-Audio 2, a latent diffusion-based T2A method that builds on the success of Make-an-Audio. Our approach includes several techniques to improve semantic alignment and temporal consistency: Firstly, we use pre-trained large language models (LLMs) to parse the text into structured <event & order> pairs for better temporal information capture. We also introduce another structured-text encoder to aid in learning semantic alignment during the diffusion denoising process. To improve the performance of variable length generation and enhance the temporal information extraction, we design a feed-forward Transformer-based diffusion denoiser. Finally, we use LLMs to augment and transform a large amount of audio-label data into audio-text datasets to alleviate the problem of scarcity of temporal data. Extensive experiments show that our method outperforms baseline models in both objective and subjective metrics, and achieves significant gains in temporal information understanding, semantic consistency, and sound quality.

Text-to-music models allow users to generate nearly realistic musical audio with textual commands. However, editing music audios remains challenging due to the conflicting desiderata of performing fine-grained alterations on the audio while maintaining a simple user interface. To address this challenge, we propose Audio Prompt Adapter (or AP-Adapter), a lightweight addition to pretrained text-to-music models. We utilize AudioMAE to extract features from the input audio, and construct attention-based adapters to feedthese features into the internal layers of AudioLDM2, a diffusion-based text-to-music model. With 22M trainable parameters, AP-Adapter empowers users to harness both global (e.g., genre and timbre) and local (e.g., melody) aspects of music, using the original audio and a short text as inputs. Through objective and subjective studies, we evaluate AP-Adapter on three tasks: timbre transfer, genre transfer, and accompaniment generation. Additionally, we demonstrate its effectiveness on out-of-domain audios containing unseen instruments during training.

Foley is a term commonly used in filmmaking, referring to the addition of daily sound effects to silent films or videos to enhance the auditory experience. Video-to-Audio (V2A), as a particular type of automatic foley task, presents inherent challenges related to audio-visual synchronization. These challenges encompass maintaining the content consistency between the input video and the generated audio, as well as the alignment of temporal and loudness properties within the video. To address these issues, we construct a controllable video-to-audio synthesis model, termed Draw an Audio, which supports multiple input instructions through drawn masks and loudness signals. To ensure content consistency between the synthesized audio and target video, we introduce the Mask-Attention Module (MAM), which employs masked video instruction to enable the model to focus on regions of interest. Additionally, we implement the Time-Loudness Module (TLM), which uses an auxiliary loudness signal to ensure the synthesis of sound that aligns with the video in both loudness and temporal dimensions. Furthermore, we have extended a large-scale V2A dataset, named VGGSound-Caption, by annotating caption prompts. Extensive experiments on challenging benchmarks across two large-scale V2A datasets verify Draw an Audio achieves the state-of-the-art. Project page: https://yannqi.github.io/Draw-an-Audio/.

Recently, denoising diffusion models have demonstrated remarkable performance among generative models in various domains. However, in the speech domain, the application of diffusion models for synthesizing time-varying audio faces limitations in terms of complexity and controllability, as speech synthesis requires very high-dimensional samples with long-term acoustic features. To alleviate the challenges posed by model complexity in singing voice synthesis, we propose HiddenSinger, a high-quality singing voice synthesis system using a neural audio codec and latent diffusion models. To ensure high-fidelity audio, we introduce an audio autoencoder that can encode audio into an audio codec as a compressed representation and reconstruct the high-fidelity audio from the low-dimensional compressed latent vector. Subsequently, we use the latent diffusion models to sample a latent representation from a musical score. In addition, our proposed model is extended to an unsupervised singing voice learning framework, HiddenSinger-U, to train the model using an unlabeled singing voice dataset. Experimental results demonstrate that our model outperforms previous models in terms of audio quality. Furthermore, the HiddenSinger-U can synthesize high-quality singing voices of speakers trained solely on unlabeled data.

Large language models have proven themselves highly flexible, able to solve a wide range of generative tasks, such as abstractive summarization and open-ended question answering. In this paper we extend the capabilities of LLMs by directly attaching a small audio encoder allowing it to perform speech recognition. By directly prepending a sequence of audial embeddings to the text token embeddings, the LLM can be converted to an automatic speech recognition (ASR) system, and be used in the exact same manner as its textual counterpart. Experiments on Multilingual LibriSpeech (MLS) show that incorporating a conformer encoder into the open sourced LLaMA-7B allows it to outperform monolingual baselines by 18% and perform multilingual speech recognition despite LLaMA being trained overwhelmingly on English text. Furthermore, we perform ablation studies to investigate whether the LLM can be completely frozen during training to maintain its original capabilities, scaling up the audio encoder, and increasing the audio encoder striding to generate fewer embeddings. The results from these studies show that multilingual ASR is possible even when the LLM is frozen or when strides of almost 1 second are used in the audio encoder opening up the possibility for LLMs to operate on long-form audio.

Large-scale pre-trained image-text models demonstrate remarkable versatility across diverse tasks, benefiting from their robust representational capabilities and effective multimodal alignment. We extend the application of these models, specifically CLIP, to the domain of sound source localization. Unlike conventional approaches, we employ the pre-trained CLIP model without explicit text input, relying solely on the audio-visual correspondence. To this end, we introduce a framework that translates audio signals into tokens compatible with CLIP's text encoder, yielding audio-driven embeddings. By directly using these embeddings, our method generates audio-grounded masks for the provided audio, extracts audio-grounded image features from the highlighted regions, and aligns them with the audio-driven embeddings using the audio-visual correspondence objective. Our findings suggest that utilizing pre-trained image-text models enable our model to generate more complete and compact localization maps for the sounding objects. Extensive experiments show that our method outperforms state-of-the-art approaches by a significant margin.

The immense scale of the recent large language models (LLM) allows many interesting properties, such as, instruction- and chain-of-thought-based fine-tuning, that has significantly improved zero- and few-shot performance in many natural language processing (NLP) tasks. Inspired by such successes, we adopt such an instruction-tuned LLM Flan-T5 as the text encoder for text-to-audio (TTA) generation -- a task where the goal is to generate an audio from its textual description. The prior works on TTA either pre-trained a joint text-audio encoder or used a non-instruction-tuned model, such as, T5. Consequently, our latent diffusion model (LDM)-based approach TANGO outperforms the state-of-the-art AudioLDM on most metrics and stays comparable on the rest on AudioCaps test set, despite training the LDM on a 63 times smaller dataset and keeping the text encoder frozen. This improvement might also be attributed to the adoption of audio pressure level-based sound mixing for training set augmentation, whereas the prior methods take a random mix.

Masked Autoencoders (MAEs) learn generalizable representations for image, text, audio, video, etc., by reconstructing masked input data from tokens of the visible data. Current MAE approaches for videos rely on random patch, tube, or frame-based masking strategies to select these tokens. This paper proposes AdaMAE, an adaptive masking strategy for MAEs that is end-to-end trainable. Our adaptive masking strategy samples visible tokens based on the semantic context using an auxiliary sampling network. This network estimates a categorical distribution over spacetime-patch tokens. The tokens that increase the expected reconstruction error are rewarded and selected as visible tokens, motivated by the policy gradient algorithm in reinforcement learning. We show that AdaMAE samples more tokens from the high spatiotemporal information regions, thereby allowing us to mask 95% of tokens, resulting in lower memory requirements and faster pre-training. We conduct ablation studies on the Something-Something v2 (SSv2) dataset to demonstrate the efficacy of our adaptive sampling approach and report state-of-the-art results of 70.0% and 81.7% in top-1 accuracy on SSv2 and Kinetics-400 action classification datasets with a ViT-Base backbone and 800 pre-training epochs.

Video recordings of speech contain correlated audio and visual information, providing a strong signal for speech representation learning from the speaker's lip movements and the produced sound. We introduce Audio-Visual Hidden Unit BERT (AV-HuBERT), a self-supervised representation learning framework for audio-visual speech, which masks multi-stream video input and predicts automatically discovered and iteratively refined multimodal hidden units. AV-HuBERT learns powerful audio-visual speech representation benefiting both lip-reading and automatic speech recognition. On the largest public lip-reading benchmark LRS3 (433 hours), AV-HuBERT achieves 32.5% WER with only 30 hours of labeled data, outperforming the former state-of-the-art approach (33.6%) trained with a thousand times more transcribed video data (31K hours). The lip-reading WER is further reduced to 26.9% when using all 433 hours of labeled data from LRS3 and combined with self-training. Using our audio-visual representation on the same benchmark for audio-only speech recognition leads to a 40% relative WER reduction over the state-of-the-art performance (1.3% vs 2.3%). Our code and models are available at https://github.com/facebookresearch/av_hubert

The ability to comprehend audio--which includes speech, non-speech sounds, and music--is crucial for AI agents to interact effectively with the world. We present MMAU, a novel benchmark designed to evaluate multimodal audio understanding models on tasks requiring expert-level knowledge and complex reasoning. MMAU comprises 10k carefully curated audio clips paired with human-annotated natural language questions and answers spanning speech, environmental sounds, and music. It includes information extraction and reasoning questions, requiring models to demonstrate 27 distinct skills across unique and challenging tasks. Unlike existing benchmarks, MMAU emphasizes advanced perception and reasoning with domain-specific knowledge, challenging models to tackle tasks akin to those faced by experts. We assess 18 open-source and proprietary (Large) Audio-Language Models, demonstrating the significant challenges posed by MMAU. Notably, even the most advanced Gemini Pro v1.5 achieves only 52.97% accuracy, and the state-of-the-art open-source Qwen2-Audio achieves only 52.50%, highlighting considerable room for improvement. We believe MMAU will drive the audio and multimodal research community to develop more advanced audio understanding models capable of solving complex audio tasks.

Lip-to-speech involves generating a natural-sounding speech synchronized with a soundless video of a person talking. Despite recent advances, current methods still cannot produce high-quality speech with high levels of intelligibility for challenging and realistic datasets such as LRS3. In this work, we present LipVoicer, a novel method that generates high-quality speech, even for in-the-wild and rich datasets, by incorporating the text modality. Given a silent video, we first predict the spoken text using a pre-trained lip-reading network. We then condition a diffusion model on the video and use the extracted text through a classifier-guidance mechanism where a pre-trained ASR serves as the classifier. LipVoicer outperforms multiple lip-to-speech baselines on LRS2 and LRS3, which are in-the-wild datasets with hundreds of unique speakers in their test set and an unrestricted vocabulary. Moreover, our experiments show that the inclusion of the text modality plays a major role in the intelligibility of the produced speech, readily perceptible while listening, and is empirically reflected in the substantial reduction of the WER metric. We demonstrate the effectiveness of LipVoicer through human evaluation, which shows that it produces more natural and synchronized speech signals compared to competing methods. Finally, we created a demo showcasing LipVoicer's superiority in producing natural, synchronized, and intelligible speech, providing additional evidence of its effectiveness. Project page and code: https://github.com/yochaiye/LipVoicer

In this work, we present CleanUNet, a causal speech denoising model on the raw waveform. The proposed model is based on an encoder-decoder architecture combined with several self-attention blocks to refine its bottleneck representations, which is crucial to obtain good results. The model is optimized through a set of losses defined over both waveform and multi-resolution spectrograms. The proposed method outperforms the state-of-the-art models in terms of denoised speech quality from various objective and subjective evaluation metrics. We release our code and models at https://github.com/nvidia/cleanunet.

Prompts are crucial to large language models as they provide context information such as topic or logical relationships. Inspired by this, we propose PromptASR, a framework that integrates prompts in end-to-end automatic speech recognition (E2E ASR) systems to achieve contextualized ASR with controllable style of transcriptions. Specifically, a dedicated text encoder encodes the text prompts and the encodings are injected into the speech encoder by cross-attending the features from two modalities. When using the ground truth text from preceding utterances as content prompt, the proposed system achieves 21.9% and 6.8% relative word error rate reductions on a book reading dataset and an in-house dataset compared to a baseline ASR system. The system can also take word-level biasing lists as prompt to improve recognition accuracy on rare words. An additional style prompt can be given to the text encoder and guide the ASR system to output different styles of transcriptions. The code is available at icefall.

The field of audio captioning has seen significant advancements in recent years, driven by the availability of large-scale audio datasets and advancements in deep learning techniques. In this technical report, we present our approach to audio captioning, focusing on the use of a pretrained speech-to-text Whisper model and pretraining on synthetic captions. We discuss our training procedures and present our experiments' results, which include model size variations, dataset mixtures, and other hyperparameters. Our findings demonstrate the impact of different training strategies on the performance of the audio captioning model. Our code and trained models are publicly available on GitHub and Hugging Face Hub.

There has been a long-standing quest for a unified audio-visual-text model to enable various multimodal understanding tasks, which mimics the listening, seeing and reading process of human beings. Humans tends to represent knowledge using two separate systems: one for representing verbal (textual) information and one for representing non-verbal (visual and auditory) information. These two systems can operate independently but can also interact with each other. Motivated by this understanding of human cognition, in this paper, we introduce CoAVT -- a novel cognition-inspired Correlated Audio-Visual-Text pre-training model to connect the three modalities. It contains a joint audio-visual encoder that learns to encode audio-visual synchronization information together with the audio and visual content for non-verbal information, and a text encoder to handle textual input for verbal information. To bridge the gap between modalities, CoAVT employs a query encoder, which contains a set of learnable query embeddings, and extracts the most informative audiovisual features of the corresponding text. Additionally, to leverage the correspondences between audio and vision with language respectively, we also establish the audio-text and visual-text bi-modal alignments upon the foundational audiovisual-text tri-modal alignment to enhance the multimodal representation learning. Finally, we jointly optimize CoAVT model with three multimodal objectives: contrastive loss, matching loss and language modeling loss. Extensive experiments show that CoAVT can learn strong multimodal correlations and be generalized to various downstream tasks. CoAVT establishes new state-of-the-art performance on text-video retrieval task on AudioCaps for both zero-shot and fine-tuning settings, audio-visual event classification and audio-visual retrieval tasks on AudioSet and VGGSound.

Codec-based text-to-speech (TTS) models have shown impressive quality with zero-shot voice cloning abilities. However, they often struggle with more expressive references or complex text inputs. We present MARS6, a robust encoder-decoder transformer for rapid, expressive TTS. MARS6 is built on recent improvements in spoken language modelling. Utilizing a hierarchical setup for its decoder, new speech tokens are processed at a rate of only 12 Hz, enabling efficient modelling of long-form text while retaining reconstruction quality. We combine several recent training and inference techniques to reduce repetitive generation and improve output stability and quality. This enables the 70M-parameter MARS6 to achieve similar performance to models many times larger. We show this in objective and subjective evaluations, comparing TTS output quality and reference speaker cloning ability. Project page: https://camb-ai.github.io/mars6-turbo/

In recent years, image generation has shown a great leap in performance, where diffusion models play a central role. Although generating high-quality images, such models are mainly conditioned on textual descriptions. This begs the question: "how can we adopt such models to be conditioned on other modalities?". In this paper, we propose a novel method utilizing latent diffusion models trained for text-to-image-generation to generate images conditioned on audio recordings. Using a pre-trained audio encoding model, the proposed method encodes audio into a new token, which can be considered as an adaptation layer between the audio and text representations. Such a modeling paradigm requires a small number of trainable parameters, making the proposed approach appealing for lightweight optimization. Results suggest the proposed method is superior to the evaluated baseline methods, considering objective and subjective metrics. Code and samples are available at: https://pages.cs.huji.ac.il/adiyoss-lab/AudioToken.

Recent advancements in audio generation have been significantly propelled by the capabilities of Large Language Models (LLMs). The existing research on audio LLM has primarily focused on enhancing the architecture and scale of audio language models, as well as leveraging larger datasets, and generally, acoustic codecs, such as EnCodec, are used for audio tokenization. However, these codecs were originally designed for audio compression, which may lead to suboptimal performance in the context of audio LLM. Our research aims to address the shortcomings of current audio LLM codecs, particularly their challenges in maintaining semantic integrity in generated audio. For instance, existing methods like VALL-E, which condition acoustic token generation on text transcriptions, often suffer from content inaccuracies and elevated word error rates (WER) due to semantic misinterpretations of acoustic tokens, resulting in word skipping and errors. To overcome these issues, we propose a straightforward yet effective approach called X-Codec. X-Codec incorporates semantic features from a pre-trained semantic encoder before the Residual Vector Quantization (RVQ) stage and introduces a semantic reconstruction loss after RVQ. By enhancing the semantic ability of the codec, X-Codec significantly reduces WER in speech synthesis tasks and extends these benefits to non-speech applications, including music and sound generation. Our experiments in text-to-speech, music continuation, and text-to-sound tasks demonstrate that integrating semantic information substantially improves the overall performance of language models in audio generation. Our code and demo are available (Demo: https://x-codec-audio.github.io Code: https://github.com/zhenye234/xcodec)

The virtual world is being established in which digital humans are created indistinguishable from real humans. Producing their audio-related capabilities is crucial since voice conveys extensive personal characteristics. We aim to create a controllable audio-form virtual singer; however, supervised modeling and controlling all different factors of the singing voice, such as timbre, tempo, pitch, and lyrics, is extremely difficult since accurately labeling all such information needs enormous labor work. In this paper, we propose a framework that could digitize a person's voice by simply "listening" to the clean voice recordings of any content in a fully unsupervised manner and predict singing voices even only using speaking recordings. A variational auto-encoder (VAE) based framework is developed, which leverages a set of pre-trained models to encode the audio as various hidden embeddings representing different factors of the singing voice, and further decodes the embeddings into raw audio. By manipulating the hidden embeddings for different factors, the resulting singing voices can be controlled, and new virtual singers can also be further generated by interpolating between timbres. Evaluations of different types of experiments demonstrate the proposed method's effectiveness. The proposed method is the critical technique for producing the AI choir, which empowered the human-AI symbiotic orchestra in Hong Kong in July 2022.

Large-scale audio language models (ALMs), such as Qwen2-Audio, are capable of comprehending diverse audio signal, performing audio analysis and generating textual responses. However, in speech emotion recognition (SER), ALMs often suffer from hallucinations, resulting in misclassifications or irrelevant outputs. To address these challenges, we propose C^2SER, a novel ALM designed to enhance the stability and accuracy of SER through Contextual perception and Chain of Thought (CoT). C^2SER integrates the Whisper encoder for semantic perception and Emotion2Vec-S for acoustic perception, where Emotion2Vec-S extends Emotion2Vec with semi-supervised learning to enhance emotional discrimination. Additionally, C^2SER employs a CoT approach, processing SER in a step-by-step manner while leveraging speech content and speaking styles to improve recognition. To further enhance stability, C^2SER introduces self-distillation from explicit CoT to implicit CoT, mitigating error accumulation and boosting recognition accuracy. Extensive experiments show that C^2SER outperforms existing popular ALMs, such as Qwen2-Audio and SECap, delivering more stable and precise emotion recognition. We release the training code, checkpoints, and test sets to facilitate further research.

Despite the growing interest in unsupervised learning, extracting meaningful knowledge from unlabelled audio remains an open challenge. To take a step in this direction, we recently proposed a problem-agnostic speech encoder (PASE), that combines a convolutional encoder followed by multiple neural networks, called workers, tasked to solve self-supervised problems (i.e., ones that do not require manual annotations as ground truth). PASE was shown to capture relevant speech information, including speaker voice-print and phonemes. This paper proposes PASE+, an improved version of PASE for robust speech recognition in noisy and reverberant environments. To this end, we employ an online speech distortion module, that contaminates the input signals with a variety of random disturbances. We then propose a revised encoder that better learns short- and long-term speech dynamics with an efficient combination of recurrent and convolutional networks. Finally, we refine the set of workers used in self-supervision to encourage better cooperation. Results on TIMIT, DIRHA and CHiME-5 show that PASE+ significantly outperforms both the previous version of PASE as well as common acoustic features. Interestingly, PASE+ learns transferable representations suitable for highly mismatched acoustic conditions.

The pre-trained speech encoder wav2vec 2.0 performs very well on various spoken language understanding (SLU) tasks. However, on many tasks, it trails behind text encoders with textual input. To improve the understanding capability of SLU encoders, various studies have used knowledge distillation to transfer knowledge from natural language understanding (NLU) encoders. We use a very simple method of distilling from a textual sentence embedder directly into wav2vec 2.0 as pre-training, utilizing paired audio-text datasets. We observed that this method is indeed capable of improving SLU task performance in fine-tuned settings, as well as full-data and few-shot transfer on a frozen encoder. However, the model performs worse on certain tasks highlighting the strengths and weaknesses of our approach.

Latent diffusion models have shown promising results in text-to-audio (T2A) generation tasks, yet previous models have encountered difficulties in generation quality, computational cost, diffusion sampling, and data preparation. In this paper, we introduce EzAudio, a transformer-based T2A diffusion model, to handle these challenges. Our approach includes several key innovations: (1) We build the T2A model on the latent space of a 1D waveform Variational Autoencoder (VAE), avoiding the complexities of handling 2D spectrogram representations and using an additional neural vocoder. (2) We design an optimized diffusion transformer architecture specifically tailored for audio latent representations and diffusion modeling, which enhances convergence speed, training stability, and memory usage, making the training process easier and more efficient. (3) To tackle data scarcity, we adopt a data-efficient training strategy that leverages unlabeled data for learning acoustic dependencies, audio caption data annotated by audio-language models for text-to-audio alignment learning, and human-labeled data for fine-tuning. (4) We introduce a classifier-free guidance (CFG) rescaling method that simplifies EzAudio by achieving strong prompt alignment while preserving great audio quality when using larger CFG scores, eliminating the need to struggle with finding the optimal CFG score to balance this trade-off. EzAudio surpasses existing open-source models in both objective metrics and subjective evaluations, delivering realistic listening experiences while maintaining a streamlined model structure, low training costs, and an easy-to-follow training pipeline. Code, data, and pre-trained models are released at: https://haidog-yaqub.github.io/EzAudio-Page/.

Our research introduces an innovative framework for video-to-audio synthesis, which solves the problems of audio-video desynchronization and semantic loss in the audio. By incorporating a semantic alignment adapter and a temporal synchronization adapter, our method significantly improves semantic integrity and the precision of beat point synchronization, particularly in fast-paced action sequences. Utilizing a contrastive audio-visual pre-trained encoder, our model is trained with video and high-quality audio data, improving the quality of the generated audio. This dual-adapter approach empowers users with enhanced control over audio semantics and beat effects, allowing the adjustment of the controller to achieve better results. Extensive experiments substantiate the effectiveness of our framework in achieving seamless audio-visual alignment.

Generating semantically and temporally aligned audio content in accordance with video input has become a focal point for researchers, particularly following the remarkable breakthrough in text-to-video generation. In this work, we aim to offer insights into the video-to-audio generation paradigm, focusing on three crucial aspects: vision encoders, auxiliary embeddings, and data augmentation techniques. Beginning with a foundational model VTA-LDM built on a simple yet surprisingly effective intuition, we explore various vision encoders and auxiliary embeddings through ablation studies. Employing a comprehensive evaluation pipeline that emphasizes generation quality and video-audio synchronization alignment, we demonstrate that our model exhibits state-of-the-art video-to-audio generation capabilities. Furthermore, we provide critical insights into the impact of different data augmentation methods on enhancing the generation framework's overall capacity. We showcase possibilities to advance the challenge of generating synchronized audio from semantic and temporal perspectives. We hope these insights will serve as a stepping stone toward developing more realistic and accurate audio-visual generation models.

This paper describes CMU's submission to the IWSLT 2024 Simultaneous Speech Translation (SST) task for translating English speech to German text in a streaming manner. Our end-to-end speech-to-text (ST) system integrates the WavLM speech encoder, a modality adapter, and the Llama2-7B-Base model as the decoder. We employ a two-stage training approach: initially, we align the representations of speech and text, followed by full fine-tuning. Both stages are trained on MuST-c v2 data with cross-entropy loss. We adapt our offline ST model for SST using a simple fixed hold-n policy. Experiments show that our model obtains an offline BLEU score of 31.1 and a BLEU score of 29.5 under 2 seconds latency on the MuST-C-v2 tst-COMMON.

Attention-based models are appealing for multimodal processing because inputs from multiple modalities can be concatenated and fed to a single backbone network - thus requiring very little fusion engineering. The resulting representations are however fully entangled throughout the network, which may not always be desirable: in learning, contrastive audio-visual self-supervised learning requires independent audio and visual features to operate, otherwise learning collapses; in inference, evaluation of audio-visual models should be possible on benchmarks having just audio or just video. In this paper, we introduce Zorro, a technique that uses masks to control how inputs from each modality are routed inside Transformers, keeping some parts of the representation modality-pure. We apply this technique to three popular transformer-based architectures (ViT, Swin and HiP) and show that with contrastive pre-training Zorro achieves state-of-the-art results on most relevant benchmarks for multimodal tasks (AudioSet and VGGSound). Furthermore, the resulting models are able to perform unimodal inference on both video and audio benchmarks such as Kinetics-400 or ESC-50.

Language models have been effectively applied to modeling natural signals, such as images, video, speech, and audio. A crucial component of these models is the codec tokenizer, which compresses high-dimensional natural signals into lower-dimensional discrete tokens. In this paper, we introduce WavTokenizer, which offers several advantages over previous SOTA acoustic codec models in the audio domain: 1)extreme compression. By compressing the layers of quantizers and the temporal dimension of the discrete codec, one-second audio of 24kHz sampling rate requires only a single quantizer with 40 or 75 tokens. 2)improved subjective quality. Despite the reduced number of tokens, WavTokenizer achieves state-of-the-art reconstruction quality with outstanding UTMOS scores and inherently contains richer semantic information. Specifically, we achieve these results by designing a broader VQ space, extended contextual windows, and improved attention networks, as well as introducing a powerful multi-scale discriminator and an inverse Fourier transform structure. We conducted extensive reconstruction experiments in the domains of speech, audio, and music. WavTokenizer exhibited strong performance across various objective and subjective metrics compared to state-of-the-art models. We also tested semantic information, VQ utilization, and adaptability to generative models. Comprehensive ablation studies confirm the necessity of each module in WavTokenizer. The related code, demos, and pre-trained models are available at https://github.com/jishengpeng/WavTokenizer.

In this article, we adapted five recent SSL methods to the task of audio classification. The first two methods, namely Deep Co-Training (DCT) and Mean Teacher (MT), involve two collaborative neural networks. The three other algorithms, called MixMatch (MM), ReMixMatch (RMM), and FixMatch (FM), are single-model methods that rely primarily on data augmentation strategies. Using the Wide-ResNet-28-2 architecture in all our experiments, 10% of labeled data and the remaining 90% as unlabeled data for training, we first compare the error rates of the five methods on three standard benchmark audio datasets: Environmental Sound Classification (ESC-10), UrbanSound8K (UBS8K), and Google Speech Commands (GSC). In all but one cases, MM, RMM, and FM outperformed MT and DCT significantly, MM and RMM being the best methods in most experiments. On UBS8K and GSC, MM achieved 18.02% and 3.25% error rate (ER), respectively, outperforming models trained with 100% of the available labeled data, which reached 23.29% and 4.94%, respectively. RMM achieved the best results on ESC-10 (12.00% ER), followed by FM which reached 13.33%. Second, we explored adding the mixup augmentation, used in MM and RMM, to DCT, MT, and FM. In almost all cases, mixup brought consistent gains. For instance, on GSC, FM reached 4.44% and 3.31% ER without and with mixup. Our PyTorch code will be made available upon paper acceptance at https:// github. com/ Labbe ti/ SSLH.

Autonomous soundscape augmentation systems typically use trained models to pick optimal maskers to effect a desired perceptual change. While acoustic information is paramount to such systems, contextual information, including participant demographics and the visual environment, also influences acoustic perception. Hence, we propose modular modifications to an existing attention-based deep neural network, to allow early, mid-level, and late feature fusion of participant-linked, visual, and acoustic features. Ablation studies on module configurations and corresponding fusion methods using the ARAUS dataset show that contextual features improve the model performance in a statistically significant manner on the normalized ISO Pleasantness, to a mean squared error of 0.1194pm0.0012 for the best-performing all-modality model, against 0.1217pm0.0009 for the audio-only model. Soundscape augmentation systems can thereby leverage multimodal inputs for improved performance. We also investigate the impact of individual participant-linked factors using trained models to illustrate improvements in model explainability.

Neural speech codecs have gained great attention for their outstanding reconstruction with discrete token representations. It is a crucial component in generative tasks such as speech coding and large language models (LLM). However, most works based on residual vector quantization perform worse with fewer tokens due to low coding efficiency for modeling complex coupled information. In this paper, we propose a neural speech codec named FreeCodec which employs a more effective encoding framework by decomposing intrinsic properties of speech into different components: 1) a global vector is extracted as the timbre information, 2) a prosody encoder with a long stride level is used to model the prosody information, 3) the content information is from a content encoder. Using different training strategies, FreeCodec achieves state-of-the-art performance in reconstruction and disentanglement scenarios. Results from subjective and objective experiments demonstrate that our framework outperforms existing methods.

Transformer-based speech self-supervised learning (SSL) models, such as HuBERT, show surprising performance in various speech processing tasks. However, huge number of parameters in speech SSL models necessitate the compression to a more compact model for wider usage in academia or small companies. In this study, we suggest to reuse attention maps across the Transformer layers, so as to remove key and query parameters while retaining the number of layers. Furthermore, we propose a novel masking distillation strategy to improve the student model's speech representation quality. We extend the distillation loss to utilize both masked and unmasked speech frames to fully leverage the teacher model's high-quality representation. Our universal compression strategy yields the student model that achieves phoneme error rate (PER) of 7.72% and word error rate (WER) of 9.96% on the SUPERB benchmark.

In this paper, we proposed Adapitch, a multi-speaker TTS method that makes adaptation of the supervised module with untranscribed data. We design two self supervised modules to train the text encoder and mel decoder separately with untranscribed data to enhance the representation of text and mel. To better handle the prosody information in a synthesized voice, a supervised TTS module is designed conditioned on content disentangling of pitch, text, and speaker. The training phase was separated into two parts, pretrained and fixed the text encoder and mel decoder with unsupervised mode, then the supervised mode on the disentanglement of TTS. Experiment results show that the Adaptich achieved much better quality than baseline methods.

Transducer and Attention based Encoder-Decoder (AED) are two widely used frameworks for speech-to-text tasks. They are designed for different purposes and each has its own benefits and drawbacks for speech-to-text tasks. In order to leverage strengths of both modeling methods, we propose a solution by combining Transducer and Attention based Encoder-Decoder (TAED) for speech-to-text tasks. The new method leverages AED's strength in non-monotonic sequence to sequence learning while retaining Transducer's streaming property. In the proposed framework, Transducer and AED share the same speech encoder. The predictor in Transducer is replaced by the decoder in the AED model, and the outputs of the decoder are conditioned on the speech inputs instead of outputs from an unconditioned language model. The proposed solution ensures that the model is optimized by covering all possible read/write scenarios and creates a matched environment for streaming applications. We evaluate the proposed approach on the MuST-C dataset and the findings demonstrate that TAED performs significantly better than Transducer for offline automatic speech recognition (ASR) and speech-to-text translation (ST) tasks. In the streaming case, TAED outperforms Transducer in the ASR task and one ST direction while comparable results are achieved in another translation direction.

Recent advancements in language models (LMs) have demonstrated strong capabilities in semantic understanding and contextual modeling, which have flourished in generative speech enhancement (SE). However, many LM-based SE approaches primarily focus on semantic information, often neglecting the critical role of acoustic information, which leads to acoustic inconsistency after enhancement and limited generalization across diverse SE tasks. In this paper, we introduce LLaSE-G1, a LLaMA-based language model that incentivizes generalization capabilities for speech enhancement. LLaSE-G1 offers the following key contributions: First, to mitigate acoustic inconsistency, LLaSE-G1 employs continuous representations from WavLM as input and predicts speech tokens from X-Codec2, maximizing acoustic preservation. Second, to promote generalization capability, LLaSE-G1 introduces dual-channel inputs and outputs, unifying multiple SE tasks without requiring task-specific IDs. Third, LLaSE-G1 outperforms prior task-specific discriminative and generative SE models, demonstrating scaling effects at test time and emerging capabilities for unseen SE tasks. Additionally, we release our code and models to support further research in this area.

In this work, we propose to study the performance of a model trained with a sentence embedding regression loss component for the Automated Audio Captioning task. This task aims to build systems that can describe audio content with a single sentence written in natural language. Most systems are trained with the standard Cross-Entropy loss, which does not take into account the semantic closeness of the sentence. We found that adding a sentence embedding loss term reduces overfitting, but also increased SPIDEr from 0.397 to 0.418 in our first setting on the AudioCaps corpus. When we increased the weight decay value, we found our model to be much closer to the current state-of-the-art methods, with a SPIDEr score up to 0.444 compared to a 0.475 score. Moreover, this model uses eight times less trainable parameters. In this training setting, the sentence embedding loss has no more impact on the model performance.

Audio and visual signals typically occur simultaneously, and humans possess an innate ability to correlate and synchronize information from these two modalities. Recently, a challenging problem known as Audio-Visual Segmentation (AVS) has emerged, intending to produce segmentation maps for sounding objects within a scene. However, the methods proposed so far have not sufficiently integrated audio and visual information, and the computational costs have been extremely high. Additionally, the outputs of different stages have not been fully utilized. To facilitate this research, we introduce a novel Progressive Confident Masking Attention Network (PMCANet). It leverages attention mechanisms to uncover the intrinsic correlations between audio signals and visual frames. Furthermore, we design an efficient and effective cross-attention module to enhance semantic perception by selecting query tokens. This selection is determined through confidence-driven units based on the network's multi-stage predictive outputs. Experiments demonstrate that our network outperforms other AVS methods while requiring less computational resources. The code is available at: https://github.com/PrettyPlate/PCMANet.

Attention encoder-decoder model architecture is the backbone of several recent top performing foundation speech models: Whisper, Seamless, OWSM, and Canary-1B. However, the reported data and compute requirements for their training are prohibitive for many in the research community. In this work, we focus on the efficiency angle and ask the questions of whether we are training these speech models efficiently, and what can we do to improve? We argue that a major, if not the most severe, detrimental factor for training efficiency is related to the sampling strategy of sequential data. We show that negligence in mini-batch sampling leads to more than 50% computation being spent on padding. To that end, we study, profile, and optimize Canary-1B training to show gradual improvement in GPU utilization leading up to 5x increase in average batch sizes versus its original training settings. This in turn allows us to train an equivalent model using 4x less GPUs in the same wall time, or leverage the original resources and train it in 2x shorter wall time. Finally, we observe that the major inference bottleneck lies in the autoregressive decoder steps. We find that adjusting the model architecture to transfer model parameters from the decoder to the encoder results in a 3x inference speedup as measured by inverse real-time factor (RTFx) while preserving the accuracy and compute requirements for convergence. The training code and models will be available as open-source.

Audio-Language models jointly learn multimodal text and audio representations that enable Zero-Shot inference. Models rely on the encoders to create powerful representations of the input and generalize to multiple tasks ranging from sounds, music, and speech. Although models have achieved remarkable performance, there is still a performance gap with task-specific models. In this paper, we propose a Contrastive Language-Audio Pretraining model that is pretrained with a diverse collection of 4.6M audio-text pairs employing two innovative encoders for Zero-Shot inference. To learn audio representations, we trained an audio encoder on 22 audio tasks, instead of the standard training of sound event classification. To learn language representations, we trained an autoregressive decoder-only model instead of the standard encoder-only models. Then, the audio and language representations are brought into a joint multimodal space using Contrastive Learning. We used our encoders to improve the downstream performance by a margin. We extensively evaluated the generalization of our representations on 26 downstream tasks, the largest in the literature. Our model achieves state of the art results in several tasks leading the way towards general-purpose audio representations.

Generative multimodal content is increasingly prevalent in much of the content creation arena, as it has the potential to allow artists and media personnel to create pre-production mockups by quickly bringing their ideas to life. The generation of audio from text prompts is an important aspect of such processes in the music and film industry. Many of the recent diffusion-based text-to-audio models focus on training increasingly sophisticated diffusion models on a large set of datasets of prompt-audio pairs. These models do not explicitly focus on the presence of concepts or events and their temporal ordering in the output audio with respect to the input prompt. Our hypothesis is focusing on how these aspects of audio generation could improve audio generation performance in the presence of limited data. As such, in this work, using an existing text-to-audio model Tango, we synthetically create a preference dataset where each prompt has a winner audio output and some loser audio outputs for the diffusion model to learn from. The loser outputs, in theory, have some concepts from the prompt missing or in an incorrect order. We fine-tune the publicly available Tango text-to-audio model using diffusion-DPO (direct preference optimization) loss on our preference dataset and show that it leads to improved audio output over Tango and AudioLDM2, in terms of both automatic- and manual-evaluation metrics.

We tackle the problem of generating audio samples conditioned on descriptive text captions. In this work, we propose AaudioGen, an auto-regressive generative model that generates audio samples conditioned on text inputs. AudioGen operates on a learnt discrete audio representation. The task of text-to-audio generation poses multiple challenges. Due to the way audio travels through a medium, differentiating ``objects'' can be a difficult task (e.g., separating multiple people simultaneously speaking). This is further complicated by real-world recording conditions (e.g., background noise, reverberation, etc.). Scarce text annotations impose another constraint, limiting the ability to scale models. Finally, modeling high-fidelity audio requires encoding audio at high sampling rate, leading to extremely long sequences. To alleviate the aforementioned challenges we propose an augmentation technique that mixes different audio samples, driving the model to internally learn to separate multiple sources. We curated 10 datasets containing different types of audio and text annotations to handle the scarcity of text-audio data points. For faster inference, we explore the use of multi-stream modeling, allowing the use of shorter sequences while maintaining a similar bitrate and perceptual quality. We apply classifier-free guidance to improve adherence to text. Comparing to the evaluated baselines, AudioGen outperforms over both objective and subjective metrics. Finally, we explore the ability of the proposed method to generate audio continuation conditionally and unconditionally. Samples: https://felixkreuk.github.io/audiogen

Recognizing whispered speech and converting it to normal speech creates many possibilities for speech interaction. Because the sound pressure of whispered speech is significantly lower than that of normal speech, it can be used as a semi-silent speech interaction in public places without being audible to others. Converting whispers to normal speech also improves the speech quality for people with speech or hearing impairments. However, conventional speech conversion techniques do not provide sufficient conversion quality or require speaker-dependent datasets consisting of pairs of whispered and normal speech utterances. To address these problems, we propose WESPER, a zero-shot, real-time whisper-to-normal speech conversion mechanism based on self-supervised learning. WESPER consists of a speech-to-unit (STU) encoder, which generates hidden speech units common to both whispered and normal speech, and a unit-to-speech (UTS) decoder, which reconstructs speech from the encoded speech units. Unlike the existing methods, this conversion is user-independent and does not require a paired dataset for whispered and normal speech. The UTS decoder can reconstruct speech in any target speaker's voice from speech units, and it requires only an unlabeled target speaker's speech data. We confirmed that the quality of the speech converted from a whisper was improved while preserving its natural prosody. Additionally, we confirmed the effectiveness of the proposed approach to perform speech reconstruction for people with speech or hearing disabilities. (project page: http://lab.rekimoto.org/projects/wesper )

Speech enhancement systems are typically trained using pairs of clean and noisy speech. In audio-visual speech enhancement (AVSE), there is not as much ground-truth clean data available; most audio-visual datasets are collected in real-world environments with background noise and reverberation, hampering the development of AVSE. In this work, we introduce AV2Wav, a resynthesis-based audio-visual speech enhancement approach that can generate clean speech despite the challenges of real-world training data. We obtain a subset of nearly clean speech from an audio-visual corpus using a neural quality estimator, and then train a diffusion model on this subset to generate waveforms conditioned on continuous speech representations from AV-HuBERT with noise-robust training. We use continuous rather than discrete representations to retain prosody and speaker information. With this vocoding task alone, the model can perform speech enhancement better than a masking-based baseline. We further fine-tune the diffusion model on clean/noisy utterance pairs to improve the performance. Our approach outperforms a masking-based baseline in terms of both automatic metrics and a human listening test and is close in quality to the target speech in the listening test. Audio samples can be found at https://home.ttic.edu/~jcchou/demo/avse/avse_demo.html.

This paper examines the integration of real-time talking-head generation for interviewer training, focusing on overcoming challenges in Audio Feature Extraction (AFE), which often introduces latency and limits responsiveness in real-time applications. To address these issues, we propose and implement a fully integrated system that replaces conventional AFE models with Open AI's Whisper, leveraging its encoder to optimize processing and improve overall system efficiency. Our evaluation of two open-source real-time models across three different datasets shows that Whisper not only accelerates processing but also improves specific aspects of rendering quality, resulting in more realistic and responsive talking-head interactions. These advancements make the system a more effective tool for immersive, interactive training applications, expanding the potential of AI-driven avatars in interviewer training.

We study speech intent classification and slot filling (SICSF) by proposing to use an encoder pretrained on speech recognition (ASR) to initialize an end-to-end (E2E) Conformer-Transformer model, which achieves the new state-of-the-art results on the SLURP dataset, with 90.14% intent accuracy and 82.27% SLURP-F1. We compare our model with encoders pretrained on self-supervised learning (SSL), and show that ASR pretraining is much more effective than SSL for SICSF. To explore parameter efficiency, we freeze the encoder and add Adapter modules, and show that parameter efficiency is only achievable with an ASR-pretrained encoder, while the SSL encoder needs full finetuning to achieve comparable results. In addition, we provide an in-depth comparison on end-to-end models versus cascading models (ASR+NLU), and show that E2E models are better than cascaded models unless an oracle ASR model is provided. Last but not least, our model is the first E2E model that achieves the same performance as cascading models with oracle ASR. Code, checkpoints and configs are available.

Recent end-to-end automatic speech recognition (ASR) systems often utilize a Transformer-based acoustic encoder that generates embedding at a high frame rate. However, this design is inefficient, particularly for long speech signals due to the quadratic computation of self-attention. To address this, we propose a new method, Adjacent Token Merging (A-ToMe), which gradually combines adjacent tokens with high similarity scores between their key values. In this way, the total time step could be reduced, and the inference of both the encoder and joint network is accelerated. Experiments on LibriSpeech show that our method can reduce 57% of tokens and improve the inference speed on GPU by 70% without any notable loss of accuracy. Additionally, we demonstrate that A-ToMe is also an effective solution to reduce tokens in long-form ASR, where the input speech consists of multiple utterances.

Using unsupervised learning to disentangle speech into content, rhythm, pitch, and timbre for voice conversion has become a hot research topic. Existing works generally take into account disentangling speech components through human-crafted bottleneck features which can not achieve sufficient information disentangling, while pitch and rhythm may still be mixed together. There is a risk of information overlap in the disentangling process which results in less speech naturalness. To overcome such limits, we propose a two-stage model to disentangle speech representations in a self-supervised manner without a human-crafted bottleneck design, which uses the Mutual Information (MI) with the designed upper bound estimator (IFUB) to separate overlapping information between speech components. Moreover, we design a Joint Text-Guided Consistent (TGC) module to guide the extraction of speech content and eliminate timbre leakage issues. Experiments show that our model can achieve a better performance than the baseline, regarding disentanglement effectiveness, speech naturalness, and similarity. Audio samples can be found at https://largeaudiomodel.com/eadvc.

Research on multi-modal contrastive learning strategies for audio and text has rapidly gained interest. Contrastively trained Audio-Language Models (ALMs), such as CLAP, which establish a unified representation across audio and language modalities, have enhanced the efficacy in various subsequent tasks by providing good text aligned audio encoders and vice versa. These improvements are evident in areas like zero-shot audio classification and audio retrieval, among others. However, the ability of these models to understand natural language and temporal relations is still a largely unexplored and open field for research. In this paper, we propose to equip the multi-modal ALMs with temporal understanding without loosing their inherent prior capabilities of audio-language tasks with a temporal instillation method TeminAL. We implement a two-stage training scheme TeminAL A & B, where the model first learns to differentiate between multiple sounds in TeminAL A, followed by a phase that instills a sense of time, thereby enhancing its temporal understanding in TeminAL B. This approach results in an average performance gain of 5.28% in temporal understanding on the ESC-50 dataset, while the model remains competitive in zero-shot retrieval and classification tasks on the AudioCap/Clotho datasets. We also note the lack of proper evaluation techniques for contrastive ALMs and propose a strategy for evaluating ALMs in zero-shot settings. The general-purpose zero-shot model evaluation strategy ZSTE, is used to evaluate various prior models. ZSTE demonstrates a general strategy to evaluate all ZS contrastive models. The model trained with TeminAL successfully outperforms current models on most downstream tasks.

The Conformer has become the most popular encoder model for automatic speech recognition (ASR). It adds convolution modules to a transformer to learn both local and global dependencies. In this work we describe a faster, more memory-efficient, and better-performing transformer, called Zipformer. Modeling changes include: 1) a U-Net-like encoder structure where middle stacks operate at lower frame rates; 2) reorganized block structure with more modules, within which we re-use attention weights for efficiency; 3) a modified form of LayerNorm called BiasNorm allows us to retain some length information; 4) new activation functions SwooshR and SwooshL work better than Swish. We also propose a new optimizer, called ScaledAdam, which scales the update by each tensor's current scale to keep the relative change about the same, and also explictly learns the parameter scale. It achieves faster convergence and better performance than Adam. Extensive experiments on LibriSpeech, Aishell-1, and WenetSpeech datasets demonstrate the effectiveness of our proposed Zipformer over other state-of-the-art ASR models. Our code is publicly available at https://github.com/k2-fsa/icefall.

We propose to synthesize high-quality and synchronized audio, given video and optional text conditions, using a novel multimodal joint training framework MMAudio. In contrast to single-modality training conditioned on (limited) video data only, MMAudio is jointly trained with larger-scale, readily available text-audio data to learn to generate semantically aligned high-quality audio samples. Additionally, we improve audio-visual synchrony with a conditional synchronization module that aligns video conditions with audio latents at the frame level. Trained with a flow matching objective, MMAudio achieves new video-to-audio state-of-the-art among public models in terms of audio quality, semantic alignment, and audio-visual synchronization, while having a low inference time (1.23s to generate an 8s clip) and just 157M parameters. MMAudio also achieves surprisingly competitive performance in text-to-audio generation, showing that joint training does not hinder single-modality performance. Code and demo are available at: https://hkchengrex.github.io/MMAudio

Following the rationale of end-to-end modeling, CTC, RNN-T or encoder-decoder-attention models for automatic speech recognition (ASR) use graphemes or grapheme-based subword units based on e.g. byte-pair encoding (BPE). The mapping from pronunciation to spelling is learned completely from data. In contrast to this, classical approaches to ASR employ secondary knowledge sources in the form of phoneme lists to define phonetic output labels and pronunciation lexica. In this work, we do a systematic comparison between grapheme- and phoneme-based output labels for an encoder-decoder-attention ASR model. We investigate the use of single phonemes as well as BPE-based phoneme groups as output labels of our model. To preserve a simplified and efficient decoder design, we also extend the phoneme set by auxiliary units to be able to distinguish homophones. Experiments performed on the Switchboard 300h and LibriSpeech benchmarks show that phoneme-based modeling is competitive to grapheme-based encoder-decoder-attention modeling.

In recent years, datasets of paired audio and captions have enabled remarkable success in automatically generating descriptions for audio clips, namely Automated Audio Captioning (AAC). However, it is labor-intensive and time-consuming to collect a sufficient number of paired audio and captions. Motivated by the recent advances in Contrastive Language-Audio Pretraining (CLAP), we propose a weakly-supervised approach to train an AAC model assuming only text data and a pre-trained CLAP model, alleviating the need for paired target data. Our approach leverages the similarity between audio and text embeddings in CLAP. During training, we learn to reconstruct the text from the CLAP text embedding, and during inference, we decode using the audio embeddings. To mitigate the modality gap between the audio and text embeddings we employ strategies to bridge the gap during training and inference stages. We evaluate our proposed method on Clotho and AudioCaps datasets demonstrating its ability to achieve a relative performance of up to ~83% compared to fully supervised approaches trained with paired target data.

Can we leverage the audiovisual information already present in video to improve self-supervised representation learning? To answer this question, we study various pretraining architectures and objectives within the masked autoencoding framework, motivated by the success of similar methods in natural language and image understanding. We show that we can achieve significant improvements on audiovisual downstream classification tasks, surpassing the state-of-the-art on VGGSound and AudioSet. Furthermore, we can leverage our audiovisual pretraining scheme for multiple unimodal downstream tasks using a single audiovisual pretrained model. We additionally demonstrate the transferability of our representations, achieving state-of-the-art audiovisual results on Epic Kitchens without pretraining specifically for this dataset.

Current self-supervised learning algorithms are often modality-specific and require large amounts of computational resources. To address these issues, we increase the training efficiency of data2vec, a learning objective that generalizes across several modalities. We do not encode masked tokens, use a fast convolutional decoder and amortize the effort to build teacher representations. data2vec 2.0 benefits from the rich contextualized target representations introduced in data2vec which enable a fast self-supervised learner. Experiments on ImageNet-1K image classification show that data2vec 2.0 matches the accuracy of Masked Autoencoders in 16.4x lower pre-training time, on Librispeech speech recognition it performs as well as wav2vec 2.0 in 10.6x less time, and on GLUE natural language understanding it matches a retrained RoBERTa model in half the time. Trading some speed for accuracy results in ImageNet-1K top-1 accuracy of 86.8\% with a ViT-L model trained for 150 epochs.

Text-to-audio (TTA) system has recently gained attention for its ability to synthesize general audio based on text descriptions. However, previous studies in TTA have limited generation quality with high computational costs. In this study, we propose AudioLDM, a TTA system that is built on a latent space to learn the continuous audio representations from contrastive language-audio pretraining (CLAP) latents. The pretrained CLAP models enable us to train LDMs with audio embedding while providing text embedding as a condition during sampling. By learning the latent representations of audio signals and their compositions without modeling the cross-modal relationship, AudioLDM is advantageous in both generation quality and computational efficiency. Trained on AudioCaps with a single GPU, AudioLDM achieves state-of-the-art TTA performance measured by both objective and subjective metrics (e.g., frechet distance). Moreover, AudioLDM is the first TTA system that enables various text-guided audio manipulations (e.g., style transfer) in a zero-shot fashion. Our implementation and demos are available at https://audioldm.github.io.

Large-scale automatic speech translation systems today lack key features that help machine-mediated communication feel seamless when compared to human-to-human dialogue. In this work, we introduce a family of models that enable end-to-end expressive and multilingual translations in a streaming fashion. First, we contribute an improved version of the massively multilingual and multimodal SeamlessM4T model-SeamlessM4T v2. This newer model, incorporating an updated UnitY2 framework, was trained on more low-resource language data. SeamlessM4T v2 provides the foundation on which our next two models are initiated. SeamlessExpressive enables translation that preserves vocal styles and prosody. Compared to previous efforts in expressive speech research, our work addresses certain underexplored aspects of prosody, such as speech rate and pauses, while also preserving the style of one's voice. As for SeamlessStreaming, our model leverages the Efficient Monotonic Multihead Attention mechanism to generate low-latency target translations without waiting for complete source utterances. As the first of its kind, SeamlessStreaming enables simultaneous speech-to-speech/text translation for multiple source and target languages. To ensure that our models can be used safely and responsibly, we implemented the first known red-teaming effort for multimodal machine translation, a system for the detection and mitigation of added toxicity, a systematic evaluation of gender bias, and an inaudible localized watermarking mechanism designed to dampen the impact of deepfakes. Consequently, we bring major components from SeamlessExpressive and SeamlessStreaming together to form Seamless, the first publicly available system that unlocks expressive cross-lingual communication in real-time. The contributions to this work are publicly released and accessible at https://github.com/facebookresearch/seamless_communication

Recovering the masked speech frames is widely applied in speech representation learning. However, most of these models use random masking in the pre-training. In this work, we proposed two kinds of masking approaches: (1) speech-level masking, making the model to mask more speech segments than silence segments, (2) phoneme-level masking, forcing the model to mask the whole frames of the phoneme, instead of phoneme pieces. We pre-trained the model via these two approaches, and evaluated on two downstream tasks, phoneme classification and speaker recognition. The experiments demonstrated that the proposed masking approaches are beneficial to improve the performance of speech representation.

Semantic information refers to the meaning conveyed through words, phrases, and contextual relationships within a given linguistic structure. Humans can leverage semantic information, such as familiar linguistic patterns and contextual cues, to reconstruct incomplete or masked speech signals in noisy environments. However, existing speech enhancement (SE) approaches often overlook the rich semantic information embedded in speech, which is crucial for improving intelligibility, speaker consistency, and overall quality of enhanced speech signals. To enrich the SE model with semantic information, we employ language models as an efficient semantic learner and propose a comprehensive framework tailored for language model-based speech enhancement, called GenSE. Specifically, we approach SE as a conditional language modeling task rather than a continuous signal regression problem defined in existing works. This is achieved by tokenizing speech signals into semantic tokens using a pre-trained self-supervised model and into acoustic tokens using a custom-designed single-quantizer neural codec model. To improve the stability of language model predictions, we propose a hierarchical modeling method that decouples the generation of clean semantic tokens and clean acoustic tokens into two distinct stages. Moreover, we introduce a token chain prompting mechanism during the acoustic token generation stage to ensure timbre consistency throughout the speech enhancement process. Experimental results on benchmark datasets demonstrate that our proposed approach outperforms state-of-the-art SE systems in terms of speech quality and generalization capability.

We introduce Metis, a foundation model for unified speech generation. Unlike previous task-specific or multi-task models, Metis follows a pre-training and fine-tuning paradigm. It is pre-trained on large-scale unlabeled speech data using masked generative modeling and then fine-tuned to adapt to diverse speech generation tasks. Specifically, 1) Metis utilizes two discrete speech representations: SSL tokens derived from speech self-supervised learning (SSL) features, and acoustic tokens directly quantized from waveforms. 2) Metis performs masked generative pre-training on SSL tokens, utilizing 300K hours of diverse speech data, without any additional condition. 3) Through fine-tuning with task-specific conditions, Metis achieves efficient adaptation to various speech generation tasks while supporting multimodal input, even when using limited data and trainable parameters. Experiments demonstrate that Metis can serve as a foundation model for unified speech generation: Metis outperforms state-of-the-art task-specific or multi-task systems across five speech generation tasks, including zero-shot text-to-speech, voice conversion, target speaker extraction, speech enhancement, and lip-to-speech, even with fewer than 20M trainable parameters or 300 times less training data. Audio samples are are available at https://metis-demo.github.io/.

Automatic speech recognition systems have created exciting possibilities for applications, however they also enable opportunities for systematic eavesdropping. We propose a method to camouflage a person's voice over-the-air from these systems without inconveniencing the conversation between people in the room. Standard adversarial attacks are not effective in real-time streaming situations because the characteristics of the signal will have changed by the time the attack is executed. We introduce predictive attacks, which achieve real-time performance by forecasting the attack that will be the most effective in the future. Under real-time constraints, our method jams the established speech recognition system DeepSpeech 3.9x more than baselines as measured through word error rate, and 6.6x more as measured through character error rate. We furthermore demonstrate our approach is practically effective in realistic environments over physical distances.

This paper studies a novel pre-training technique with unpaired speech data, Speech2C, for encoder-decoder based automatic speech recognition (ASR). Within a multi-task learning framework, we introduce two pre-training tasks for the encoder-decoder network using acoustic units, i.e., pseudo codes, derived from an offline clustering model. One is to predict the pseudo codes via masked language modeling in encoder output, like HuBERT model, while the other lets the decoder learn to reconstruct pseudo codes autoregressively instead of generating textual scripts. In this way, the decoder learns to reconstruct original speech information with codes before learning to generate correct text. Comprehensive experiments on the LibriSpeech corpus show that the proposed Speech2C can relatively reduce the word error rate (WER) by 19.2% over the method without decoder pre-training, and also outperforms significantly the state-of-the-art wav2vec 2.0 and HuBERT on fine-tuning subsets of 10h and 100h. We release our code and model at https://github.com/microsoft/SpeechT5/tree/main/Speech2C.

This paper presents an innovative approach to enhance control over audio generation by emphasizing the alignment between audio and text representations during model training. In the context of language model-based audio generation, the model leverages input from both textual and audio token representations to predict subsequent audio tokens. However, the current configuration lacks explicit regularization to ensure the alignment between the chosen text representation and the language model's predictions. Our proposal involves the incorporation of audio and text representation regularization, particularly during the classifier-free guidance (CFG) phase, where the text condition is excluded from cross attention during language model training. The aim of this proposed representation regularization is to minimize discrepancies in audio and text similarity compared to other samples within the same training batch. Experimental results on both music and audio generation tasks demonstrate that our proposed methods lead to improvements in objective metrics for both audio and music generation, as well as an enhancement in the human perception for audio generation.

Discrete audio tokens have recently gained attention for their potential to bridge the gap between audio and language processing. Ideal audio tokens must preserve content, paralinguistic elements, speaker identity, and many other audio details. Current audio tokenization methods fall into two categories: Semantic tokens, acquired through quantization of Self-Supervised Learning (SSL) models, and Neural compression-based tokens (codecs). Although previous studies have benchmarked codec models to identify optimal configurations, the ideal setup for quantizing pretrained SSL models remains unclear. This paper explores the optimal configuration of semantic tokens across discriminative and generative tasks. We propose a scalable solution to train a universal vocoder across multiple SSL layers. Furthermore, an attention mechanism is employed to identify task-specific influential layers, enhancing the adaptability and performance of semantic tokens in diverse audio applications.

The last year has seen astonishing progress in text-prompted image generation premised on the idea of a cross-modal representation space in which the text and image domains are represented jointly. In ASR, this idea has found application as joint speech-text encoders that can scale to the capacities of very large parameter models by being trained on both unpaired speech and text. While these methods show promise, they have required special treatment of the sequence-length mismatch inherent in speech and text, either by up-sampling heuristics or an explicit alignment model. In this work, we offer evidence that joint speech-text encoders naturally achieve consistent representations across modalities by disregarding sequence length, and argue that consistency losses could forgive length differences and simply assume the best alignment. We show that such a loss improves downstream WER in both a large-parameter monolingual and multilingual system.

Zero-shot text-to-speech (TTS) has gained significant attention due to its powerful voice cloning capabilities, requiring only a few seconds of unseen speaker voice prompts. However, all previous work has been developed for cloud-based systems. Taking autoregressive models as an example, although these approaches achieve high-fidelity voice cloning, they fall short in terms of inference speed, model size, and robustness. Therefore, we propose MobileSpeech, which is a fast, lightweight, and robust zero-shot text-to-speech system based on mobile devices for the first time. Specifically: 1) leveraging discrete codec, we design a parallel speech mask decoder module called SMD, which incorporates hierarchical information from the speech codec and weight mechanisms across different codec layers during the generation process. Moreover, to bridge the gap between text and speech, we introduce a high-level probabilistic mask that simulates the progression of information flow from less to more during speech generation. 2) For speaker prompts, we extract fine-grained prompt duration from the prompt speech and incorporate text, prompt speech by cross attention in SMD. We demonstrate the effectiveness of MobileSpeech on multilingual datasets at different levels, achieving state-of-the-art results in terms of generating speed and speech quality. MobileSpeech achieves RTF of 0.09 on a single A100 GPU and we have successfully deployed MobileSpeech on mobile devices. Audio samples are available at https://mobilespeech.github.io/ .

Large-scale multimodal generative modeling has created milestones in text-to-image and text-to-video generation. Its application to audio still lags behind for two main reasons: the lack of large-scale datasets with high-quality text-audio pairs, and the complexity of modeling long continuous audio data. In this work, we propose Make-An-Audio with a prompt-enhanced diffusion model that addresses these gaps by 1) introducing pseudo prompt enhancement with a distill-then-reprogram approach, it alleviates data scarcity with orders of magnitude concept compositions by using language-free audios; 2) leveraging spectrogram autoencoder to predict the self-supervised audio representation instead of waveforms. Together with robust contrastive language-audio pretraining (CLAP) representations, Make-An-Audio achieves state-of-the-art results in both objective and subjective benchmark evaluation. Moreover, we present its controllability and generalization for X-to-Audio with "No Modality Left Behind", for the first time unlocking the ability to generate high-definition, high-fidelity audios given a user-defined modality input. Audio samples are available at https://Text-to-Audio.github.io

We present SoundStream, a novel neural audio codec that can efficiently compress speech, music and general audio at bitrates normally targeted by speech-tailored codecs. SoundStream relies on a model architecture composed by a fully convolutional encoder/decoder network and a residual vector quantizer, which are trained jointly end-to-end. Training leverages recent advances in text-to-speech and speech enhancement, which combine adversarial and reconstruction losses to allow the generation of high-quality audio content from quantized embeddings. By training with structured dropout applied to quantizer layers, a single model can operate across variable bitrates from 3kbps to 18kbps, with a negligible quality loss when compared with models trained at fixed bitrates. In addition, the model is amenable to a low latency implementation, which supports streamable inference and runs in real time on a smartphone CPU. In subjective evaluations using audio at 24kHz sampling rate, SoundStream at 3kbps outperforms Opus at 12kbps and approaches EVS at 9.6kbps. Moreover, we are able to perform joint compression and enhancement either at the encoder or at the decoder side with no additional latency, which we demonstrate through background noise suppression for speech.

Singing voice synthesis and conversion have emerged as significant subdomains of voice generation, leading to much demands on prompt-conditioned generation. Unlike common voice data, generating a singing voice requires an understanding of various associated vocal and musical characteristics, such as the vocal tone of the singer or emotional expressions. However, existing open-source audio-text datasets for voice generation tend to capture only a very limited range of attributes, often missing musical characteristics of the audio. To fill this gap, we introduce S2Cap, an audio-text pair dataset with a diverse set of attributes. S2Cap consists of pairs of textual prompts and music audio samples with a wide range of vocal and musical attributes, including pitch, volume, tempo, mood, singer's gender and age, and musical genre and emotional expression. Utilizing S2Cap, we suggest an effective novel baseline algorithm for singing style captioning. Singing style captioning is a relative task to voice generation that generates text descriptions of vocal characteristics, which we first suggested. First, to mitigate the misalignment between the audio encoder and the text decoder, we present a novel mechanism called CRESCENDO, which utilizes positive-pair similarity learning to synchronize the embedding spaces of a pretrained audio encoder to get similar embeddings with a text encoder. We additionally supervise the model using the singer's voice, which is demixed by the accompaniment. This supervision allows the model to more accurately capture vocal characteristics, leading to improved singing style captions that better reflect the style of the singer. The dataset and the codes are available at https://github.com/HJ-Ok/S2cap.

In this work, we present the Textless Vision-Language Transformer (TVLT), where homogeneous transformer blocks take raw visual and audio inputs for vision-and-language representation learning with minimal modality-specific design, and do not use text-specific modules such as tokenization or automatic speech recognition (ASR). TVLT is trained by reconstructing masked patches of continuous video frames and audio spectrograms (masked autoencoding) and contrastive modeling to align video and audio. TVLT attains performance comparable to its text-based counterpart on various multimodal tasks, such as visual question answering, image retrieval, video retrieval, and multimodal sentiment analysis, with 28x faster inference speed and only 1/3 of the parameters. Our findings suggest the possibility of learning compact and efficient visual-linguistic representations from low-level visual and audio signals without assuming the prior existence of text. Our code and checkpoints are available at: https://github.com/zinengtang/TVLT

Modern automatic speech recognition (ASR) models, such as OpenAI's Whisper, rely on deep encoder-decoder architectures, and their encoders are a critical bottleneck for efficient deployment due to high computational intensity. We introduce LiteASR, a low-rank compression scheme for ASR encoders that significantly reduces inference costs while maintaining transcription accuracy. Our approach leverages the strong low-rank properties observed in intermediate activations: by applying principal component analysis (PCA) with a small calibration dataset, we approximate linear transformations with a chain of low-rank matrix multiplications, and further optimize self-attention to work in the reduced dimension. Evaluation results show that our method can compress Whisper large-v3's encoder size by over 50%, matching Whisper medium's size with better transcription accuracy, thereby establishing a new Pareto-optimal frontier of efficiency and performance. The code of LiteASR is available at https://github.com/efeslab/LiteASR.

We introduce Matcha-TTS, a new encoder-decoder architecture for speedy TTS acoustic modelling, trained using optimal-transport conditional flow matching (OT-CFM). This yields an ODE-based decoder capable of high output quality in fewer synthesis steps than models trained using score matching. Careful design choices additionally ensure each synthesis step is fast to run. The method is probabilistic, non-autoregressive, and learns to speak from scratch without external alignments. Compared to strong pre-trained baseline models, the Matcha-TTS system has the smallest memory footprint, rivals the speed of the fastest models on long utterances, and attains the highest mean opinion score in a listening test. Please see https://shivammehta25.github.io/Matcha-TTS/ for audio examples, code, and pre-trained models.

Audio codec models are widely used in audio communication as a crucial technique for compressing audio into discrete representations. Nowadays, audio codec models are increasingly utilized in generation fields as intermediate representations. For instance, AudioLM is an audio generation model that uses the discrete representation of SoundStream as a training target, while VALL-E employs the Encodec model as an intermediate feature to aid TTS tasks. Despite their usefulness, two challenges persist: (1) training these audio codec models can be difficult due to the lack of publicly available training processes and the need for large-scale data and GPUs; (2) achieving good reconstruction performance requires many codebooks, which increases the burden on generation models. In this study, we propose a group-residual vector quantization (GRVQ) technique and use it to develop a novel High Fidelity Audio Codec model, HiFi-Codec, which only requires 4 codebooks. We train all the models using publicly available TTS data such as LibriTTS, VCTK, AISHELL, and more, with a total duration of over 1000 hours, using 8 GPUs. Our experimental results show that HiFi-Codec outperforms Encodec in terms of reconstruction performance despite requiring only 4 codebooks. To facilitate research in audio codec and generation, we introduce AcademiCodec, the first open-source audio codec toolkit that offers training codes and pre-trained models for Encodec, SoundStream, and HiFi-Codec. Code and pre-trained model can be found on: https://github.com/yangdongchao/AcademiCodec{https://github.com/yangdongchao/AcademiCodec}

Masked Language Modeling (MLM) has been widely used as the denoising objective in pre-training language models (PrLMs). Existing PrLMs commonly adopt a Random-Token Masking strategy where a fixed masking ratio is applied and different contents are masked by an equal probability throughout the entire training. However, the model may receive complicated impact from pre-training status, which changes accordingly as training time goes on. In this paper, we show that such time-invariant MLM settings on masking ratio and masked content are unlikely to deliver an optimal outcome, which motivates us to explore the influence of time-variant MLM settings. We propose two scheduled masking approaches that adaptively tune the masking ratio and masked content in different training stages, which improves the pre-training efficiency and effectiveness verified on the downstream tasks. Our work is a pioneer study on time-variant masking strategy on ratio and content and gives a better understanding of how masking ratio and masked content influence the MLM pre-training.

Automatic Speech Recognition (ASR) systems convert speech into text and can be placed into two broad categories: traditional and fully end-to-end. Both types have been shown to be vulnerable to adversarial audio examples that sound benign to the human ear but force the ASR to produce malicious transcriptions. Of these attacks, only the "psychoacoustic" attacks can create examples with relatively imperceptible perturbations, as they leverage the knowledge of the human auditory system. Unfortunately, existing psychoacoustic attacks can only be applied against traditional models, and are obsolete against the newer, fully end-to-end ASRs. In this paper, we propose an equalization-based psychoacoustic attack that can exploit both traditional and fully end-to-end ASRs. We successfully demonstrate our attack against real-world ASRs that include DeepSpeech and Wav2Letter. Moreover, we employ a user study to verify that our method creates low audible distortion. Specifically, 80 of the 100 participants voted in favor of all our attack audio samples as less noisier than the existing state-of-the-art attack. Through this, we demonstrate both types of existing ASR pipelines can be exploited with minimum degradation to attack audio quality.

The recent advancements in large language models (LLMs) have revolutionized the field of natural language processing, progressively broadening their scope to multimodal perception and generation. However, effectively integrating listening capabilities into LLMs poses significant challenges, particularly with respect to generalizing across varied contexts and executing complex auditory tasks. In this work, we introduce WavLLM, a robust and adaptive speech large language model with dual encoders, and a prompt-aware LoRA weight adapter, optimized by a two-stage curriculum learning approach. Leveraging dual encoders, we decouple different types of speech information, utilizing a Whisper encoder to process the semantic content of speech, and a WavLM encoder to capture the unique characteristics of the speaker's identity. Within the curriculum learning framework, WavLLM first builds its foundational capabilities by optimizing on mixed elementary single tasks, followed by advanced multi-task training on more complex tasks such as combinations of the elementary tasks. To enhance the flexibility and adherence to different tasks and instructions, a prompt-aware LoRA weight adapter is introduced in the second advanced multi-task training stage. We validate the proposed model on universal speech benchmarks including tasks such as ASR, ST, SV, ER, and also apply it to specialized datasets like Gaokao English listening comprehension set for SQA, and speech Chain-of-Thought (CoT) evaluation set. Experiments demonstrate that the proposed model achieves state-of-the-art performance across a range of speech tasks on the same model size, exhibiting robust generalization capabilities in executing complex tasks using CoT approach. Furthermore, our model successfully completes Gaokao tasks without specialized training. The codes, models, audio, and Gaokao evaluation set can be accessed at aka.ms/wavllm.

Voice conversion as the style transfer task applied to speech, refers to converting one person's speech into a new speech that sounds like another person's. Up to now, there has been a lot of research devoted to better implementation of VC tasks. However, a good voice conversion model should not only match the timbre information of the target speaker, but also expressive information such as prosody, pace, pause, etc. In this context, prosody modeling is crucial for achieving expressive voice conversion that sounds natural and convincing. Unfortunately, prosody modeling is important but challenging, especially without text transcriptions. In this paper, we firstly propose a novel voice conversion framework named 'PMVC', which effectively separates and models the content, timbre, and prosodic information from the speech without text transcriptions. Specially, we introduce a new speech augmentation algorithm for robust prosody extraction. And building upon this, mask and predict mechanism is applied in the disentanglement of prosody and content information. The experimental results on the AIShell-3 corpus supports our improvement of naturalness and similarity of converted speech.

Collecting audio-text pairs is expensive; however, it is much easier to access text-only data. Unless using shallow fusion, end-to-end automatic speech recognition (ASR) models require architecture modifications or additional training schemes to use text-only data. Inspired by recent advances in decoder-only language models (LMs), such as GPT-3 and PaLM adopted for speech-processing tasks, we propose using a decoder-only architecture for ASR with simple text augmentation. To provide audio information, encoder features compressed by CTC prediction are used as prompts for the decoder, which can be regarded as refining CTC prediction using the decoder-only model. Because the decoder architecture is the same as an autoregressive LM, it is simple to enhance the model by leveraging external text data with LM training. An experimental comparison using LibriSpeech and Switchboard shows that our proposed models with text augmentation training reduced word error rates from ordinary CTC by 0.3% and 1.4% on LibriSpeech test-clean and testother set, respectively, and 2.9% and 5.0% on Switchboard and CallHome. The proposed model had advantage on computational efficiency compared with conventional encoder-decoder ASR models with a similar parameter setup, and outperformed them on the LibriSpeech 100h and Switchboard training scenarios.

CTC compressor can be an effective approach to integrate audio encoders to decoder-only models, which has gained growing interest for different speech applications. In this work, we propose a novel CTC compressor based joint speech and text training (CJST) framework for decoder-only ASR. CJST matches speech and text modalities from both directions by exploring a simple modality adaptor and several features of the CTC compressor, including sequence compression, on-the-fly forced peaky alignment and CTC class embeddings. Experimental results on the Librispeech and TED-LIUM2 corpora show that the proposed CJST achieves an effective text injection without the need of duration handling, leading to the best performance for both in-domain and cross-domain scenarios. We also provide a comprehensive study on CTC compressor, covering various compression modes, edge case handling and behavior under both clean and noisy data conditions, which reveals the most robust setting to use CTC compressor for decoder-only models.

Self-supervised learning (SSL) has recently emerged as a promising paradigm for training generalisable models on large-scale data in the fields of vision, text, and speech. Although SSL has been proven effective in speech and audio, its application to music audio has yet to be thoroughly explored. This is primarily due to the distinctive challenges associated with modelling musical knowledge, particularly its tonal and pitched characteristics of music. To address this research gap, we propose an acoustic Music undERstanding model with large-scale self-supervised Training (MERT), which incorporates teacher models to provide pseudo labels in the masked language modelling (MLM) style acoustic pre-training. In our exploration, we identified a superior combination of teacher models, which outperforms conventional speech and audio approaches in terms of performance. This combination includes an acoustic teacher based on Residual Vector Quantization - Variational AutoEncoder (RVQ-VAE) and a musical teacher based on the Constant-Q Transform (CQT). These teachers effectively guide our student model, a BERT-style transformer encoder, to better model music audio. In addition, we introduce an in-batch noise mixture augmentation to enhance the representation robustness. Furthermore, we explore a wide range of settings to overcome the instability in acoustic language model pre-training, which allows our designed paradigm to scale from 95M to 330M parameters. Experimental results indicate that our model can generalise and perform well on 14 music understanding tasks and attains state-of-the-art (SOTA) overall scores. The code and models are online: https://github.com/yizhilll/MERT.

Self-supervised learning (SSL) achieves great success in speech recognition, while limited exploration has been attempted for other speech processing tasks. As speech signal contains multi-faceted information including speaker identity, paralinguistics, spoken content, etc., learning universal representations for all speech tasks is challenging. To tackle the problem, we propose a new pre-trained model, WavLM, to solve full-stack downstream speech tasks. WavLM jointly learns masked speech prediction and denoising in pre-training. By this means, WavLM does not only keep the speech content modeling capability by the masked speech prediction, but also improves the potential to non-ASR tasks by the speech denoising. In addition, WavLM employs gated relative position bias for the Transformer structure to better capture the sequence ordering of input speech. We also scale up the training dataset from 60k hours to 94k hours. WavLM Large achieves state-of-the-art performance on the SUPERB benchmark, and brings significant improvements for various speech processing tasks on their representative benchmarks. The code and pre-trained models are available at https://aka.ms/wavlm.

Self-supervised learning (SSL) has helped extend speech technologies to more languages by reducing the need for labeled data. However, models are still far from supporting the world's 7000+ languages. We propose XEUS, a Cross-lingual Encoder for Universal Speech, trained on over 1 million hours of data across 4057 languages, extending the language coverage of SSL models 4-fold. We combine 1 million hours of speech from existing publicly accessible corpora with a newly created corpus of 7400+ hours from 4057 languages, which will be publicly released. To handle the diverse conditions of multilingual speech data, we augment the typical SSL masked prediction approach with a novel dereverberation objective, increasing robustness. We evaluate XEUS on several benchmarks, and show that it consistently outperforms or achieves comparable results to state-of-the-art (SOTA) SSL models across a variety of tasks. XEUS sets a new SOTA on the ML-SUPERB benchmark: it outperforms MMS 1B and w2v-BERT 2.0 v2 by 0.8% and 4.4% respectively, despite having less parameters or pre-training data. Checkpoints, code, and data are found in https://www.wavlab.org/activities/2024/xeus/.

Personalised speech enhancement (PSE), which extracts only the speech of a target user and removes everything else from a recorded audio clip, can potentially improve users' experiences of audio AI modules deployed in the wild. To support a large variety of downstream audio tasks, such as real-time ASR and audio-call enhancement, a PSE solution should operate in a streaming mode, i.e., input audio cleaning should happen in real-time with a small latency and real-time factor. Personalisation is typically achieved by extracting a target speaker's voice profile from an enrolment audio, in the form of a static embedding vector, and then using it to condition the output of a PSE model. However, a fixed target speaker embedding may not be optimal under all conditions. In this work, we present a streaming Transformer-based PSE model and propose a novel cross-attention approach that gives adaptive target speaker representations. We present extensive experiments and show that our proposed cross-attention approach outperforms competitive baselines consistently, even when our model is only approximately half the size.

Recent work has studied text-to-audio synthesis using large amounts of paired text-audio data. However, audio recordings with high-quality text annotations can be difficult to acquire. In this work, we approach text-to-audio synthesis using unlabeled videos and pretrained language-vision models. We propose to learn the desired text-audio correspondence by leveraging the visual modality as a bridge. We train a conditional diffusion model to generate the audio track of a video, given a video frame encoded by a pretrained contrastive language-image pretraining (CLIP) model. At test time, we first explore performing a zero-shot modality transfer and condition the diffusion model with a CLIP-encoded text query. However, we observe a noticeable performance drop with respect to image queries. To close this gap, we further adopt a pretrained diffusion prior model to generate a CLIP image embedding given a CLIP text embedding. Our results show the effectiveness of the proposed method, and that the pretrained diffusion prior can reduce the modality transfer gap. While we focus on text-to-audio synthesis, the proposed model can also generate audio from image queries, and it shows competitive performance against a state-of-the-art image-to-audio synthesis model in a subjective listening test. This study offers a new direction of approaching text-to-audio synthesis that leverages the naturally-occurring audio-visual correspondence in videos and the power of pretrained language-vision models.

Self supervised representation learning has recently attracted a lot of research interest for both the audio and visual modalities. However, most works typically focus on a particular modality or feature alone and there has been very limited work that studies the interaction between the two modalities for learning self supervised representations. We propose a framework for learning audio representations guided by the visual modality in the context of audiovisual speech. We employ a generative audio-to-video training scheme in which we animate a still image corresponding to a given audio clip and optimize the generated video to be as close as possible to the real video of the speech segment. Through this process, the audio encoder network learns useful speech representations that we evaluate on emotion recognition and speech recognition. We achieve state of the art results for emotion recognition and competitive results for speech recognition. This demonstrates the potential of visual supervision for learning audio representations as a novel way for self-supervised learning which has not been explored in the past. The proposed unsupervised audio features can leverage a virtually unlimited amount of training data of unlabelled audiovisual speech and have a large number of potentially promising applications.

Self-supervised learning (SSL) has achieved great success in speech-related tasks, driven by advancements in speech encoder architectures and the expansion of datasets. While Transformer and Conformer architectures have dominated SSL backbones, encoders like Zipformer, which excel in automatic speech recognition (ASR), remain unexplored in SSL. Concurrently, inefficiencies in data processing within existing SSL training frameworks, such as fairseq, pose challenges in managing the growing volumes of training data. To address these issues, we propose k2SSL, an open-source framework that offers faster, more memory-efficient, and better-performing self-supervised speech representation learning, with a focus on downstream ASR tasks. The optimized HuBERT and proposed Zipformer-based SSL systems exhibit substantial reductions in both training time and memory usage during SSL training. Experiments on LibriSpeech and Libri-Light demonstrate that Zipformer-based SSL systems significantly outperform comparable HuBERT and WavLM systems, achieving a relative WER reduction on dev-other/test-other of up to 34.8%/32.4% compared to HuBERT Base after supervised fine-tuning, along with a 3.5x pre-training speedup in total GPU hours.

We introduce AudioLM, a framework for high-quality audio generation with long-term consistency. AudioLM maps the input audio to a sequence of discrete tokens and casts audio generation as a language modeling task in this representation space. We show how existing audio tokenizers provide different trade-offs between reconstruction quality and long-term structure, and we propose a hybrid tokenization scheme to achieve both objectives. Namely, we leverage the discretized activations of a masked language model pre-trained on audio to capture long-term structure and the discrete codes produced by a neural audio codec to achieve high-quality synthesis. By training on large corpora of raw audio waveforms, AudioLM learns to generate natural and coherent continuations given short prompts. When trained on speech, and without any transcript or annotation, AudioLM generates syntactically and semantically plausible speech continuations while also maintaining speaker identity and prosody for unseen speakers. Furthermore, we demonstrate how our approach extends beyond speech by generating coherent piano music continuations, despite being trained without any symbolic representation of music.

This paper introduces Stereo-Talker, a novel one-shot audio-driven human video synthesis system that generates 3D talking videos with precise lip synchronization, expressive body gestures, temporally consistent photo-realistic quality, and continuous viewpoint control. The process follows a two-stage approach. In the first stage, the system maps audio input to high-fidelity motion sequences, encompassing upper-body gestures and facial expressions. To enrich motion diversity and authenticity, large language model (LLM) priors are integrated with text-aligned semantic audio features, leveraging LLMs' cross-modal generalization power to enhance motion quality. In the second stage, we improve diffusion-based video generation models by incorporating a prior-guided Mixture-of-Experts (MoE) mechanism: a view-guided MoE focuses on view-specific attributes, while a mask-guided MoE enhances region-based rendering stability. Additionally, a mask prediction module is devised to derive human masks from motion data, enhancing the stability and accuracy of masks and enabling mask guiding during inference. We also introduce a comprehensive human video dataset with 2,203 identities, covering diverse body gestures and detailed annotations, facilitating broad generalization. The code, data, and pre-trained models will be released for research purposes.

The recent advancement of speech recognition technology has been driven by large-scale datasets and attention-based architectures, but many challenges still remain, especially for low-resource languages and dialects. This paper explores the integration of weakly supervised transcripts from TV subtitles into automatic speech recognition (ASR) systems, aiming to improve both verbatim transcriptions and automatically generated subtitles. To this end, verbatim data and subtitles are regarded as different domains or languages, due to their distinct characteristics. We propose and compare several end-to-end architectures that are designed to jointly model both modalities with separate or shared encoders and decoders. The proposed methods are able to jointly generate a verbatim transcription and a subtitle. Evaluation on Flemish (Belgian Dutch) demonstrates that a model with cascaded encoders and separate decoders allows to represent the differences between the two data types most efficiently while improving on both domains. Despite differences in domain and linguistic variations, combining verbatim transcripts with subtitle data leads to notable ASR improvements without the need for extensive preprocessing. Additionally, experiments with a large-scale subtitle dataset show the scalability of the proposed approach. The methods not only improve ASR accuracy but also generate subtitles that closely match standard written text, offering several potential applications.

Recent advances in transformer-based architectures which are pre-trained in self-supervised manner have shown great promise in several machine learning tasks. In the audio domain, such architectures have also been successfully utilised in the field of speech emotion recognition (SER). However, existing works have not evaluated the influence of model size and pre-training data on downstream performance, and have shown limited attention to generalisation, robustness, fairness, and efficiency. The present contribution conducts a thorough analysis of these aspects on several pre-trained variants of wav2vec 2.0 and HuBERT that we fine-tuned on the dimensions arousal, dominance, and valence of MSP-Podcast, while additionally using IEMOCAP and MOSI to test cross-corpus generalisation. To the best of our knowledge, we obtain the top performance for valence prediction without use of explicit linguistic information, with a concordance correlation coefficient (CCC) of .638 on MSP-Podcast. Furthermore, our investigations reveal that transformer-based architectures are more robust to small perturbations compared to a CNN-based baseline and fair with respect to biological sex groups, but not towards individual speakers. Finally, we are the first to show that their extraordinary success on valence is based on implicit linguistic information learnt during fine-tuning of the transformer layers, which explains why they perform on-par with recent multimodal approaches that explicitly utilise textual information. Our findings collectively paint the following picture: transformer-based architectures constitute the new state-of-the-art in SER, but further advances are needed to mitigate remaining robustness and individual speaker issues. To make our findings reproducible, we release the best performing model to the community.

In this paper, we focus on Whisper, a recent automatic speech recognition model trained with a massive 680k hour labeled speech corpus recorded in diverse conditions. We first show an interesting finding that while Whisper is very robust against real-world background sounds (e.g., music), its audio representation is actually not noise-invariant, but is instead highly correlated to non-speech sounds, indicating that Whisper recognizes speech conditioned on the noise type. With this finding, we build a unified audio tagging and speech recognition model Whisper-AT by freezing the backbone of Whisper, and training a lightweight audio tagging model on top of it. With <1% extra computational cost, Whisper-AT can recognize audio events, in addition to spoken text, in a single forward pass.

Masked Autoencoders (MAE) have been popular paradigms for large-scale vision representation pre-training. However, MAE solely reconstructs the low-level RGB signals after the decoder and lacks supervision upon high-level semantics for the encoder, thus suffering from sub-optimal learned representations and long pre-training epochs. To alleviate this, previous methods simply replace the pixel reconstruction targets of 75% masked tokens by encoded features from pre-trained image-image (DINO) or image-language (CLIP) contrastive learning. Different from those efforts, we propose to Mimic before Reconstruct for Masked Autoencoders, named as MR-MAE, which jointly learns high-level and low-level representations without interference during pre-training. For high-level semantics, MR-MAE employs a mimic loss over 25% visible tokens from the encoder to capture the pre-trained patterns encoded in CLIP and DINO. For low-level structures, we inherit the reconstruction loss in MAE to predict RGB pixel values for 75% masked tokens after the decoder. As MR-MAE applies high-level and low-level targets respectively at different partitions, the learning conflicts between them can be naturally overcome and contribute to superior visual representations for various downstream tasks. On ImageNet-1K, the MR-MAE base pre-trained for only 400 epochs achieves 85.8% top-1 accuracy after fine-tuning, surpassing the 1600-epoch MAE base by +2.2% and the previous state-of-the-art BEiT V2 base by +0.3%. Code and pre-trained models will be released at https://github.com/Alpha-VL/ConvMAE.

This paper describes the Microsoft end-to-end neural text to speech (TTS) system: DelightfulTTS for Blizzard Challenge 2021. The goal of this challenge is to synthesize natural and high-quality speech from text, and we approach this goal in two perspectives: The first is to directly model and generate waveform in 48 kHz sampling rate, which brings higher perception quality than previous systems with 16 kHz or 24 kHz sampling rate; The second is to model the variation information in speech through a systematic design, which improves the prosody and naturalness. Specifically, for 48 kHz modeling, we predict 16 kHz mel-spectrogram in acoustic model, and propose a vocoder called HiFiNet to directly generate 48 kHz waveform from predicted 16 kHz mel-spectrogram, which can better trade off training efficiency, modelling stability and voice quality. We model variation information systematically from both explicit (speaker ID, language ID, pitch and duration) and implicit (utterance-level and phoneme-level prosody) perspectives: 1) For speaker and language ID, we use lookup embedding in training and inference; 2) For pitch and duration, we extract the values from paired text-speech data in training and use two predictors to predict the values in inference; 3) For utterance-level and phoneme-level prosody, we use two reference encoders to extract the values in training, and use two separate predictors to predict the values in inference. Additionally, we introduce an improved Conformer block to better model the local and global dependency in acoustic model. For task SH1, DelightfulTTS achieves 4.17 mean score in MOS test and 4.35 in SMOS test, which indicates the effectiveness of our proposed system

Large Language Models (LLMs) demonstrate remarkable zero-shot performance across various natural language processing tasks. The integration of multimodal encoders extends their capabilities, enabling the development of Multimodal Large Language Models that process vision, audio, and text. However, these capabilities also raise significant security concerns, as these models can be manipulated to generate harmful or inappropriate content through jailbreak. While extensive research explores the impact of modality-specific input edits on text-based LLMs and Large Vision-Language Models in jailbreak, the effects of audio-specific edits on Large Audio-Language Models (LALMs) remain underexplored. Hence, this paper addresses this gap by investigating how audio-specific edits influence LALMs inference regarding jailbreak. We introduce the Audio Editing Toolbox (AET), which enables audio-modality edits such as tone adjustment, word emphasis, and noise injection, and the Edited Audio Datasets (EADs), a comprehensive audio jailbreak benchmark. We also conduct extensive evaluations of state-of-the-art LALMs to assess their robustness under different audio edits. This work lays the groundwork for future explorations on audio-modality interactions in LALMs security.

Masked Language Modeling (MLM) has been one of the most prominent approaches for pretraining bidirectional text encoders due to its simplicity and effectiveness. One notable concern about MLM is that the special [MASK] symbol causes a discrepancy between pretraining data and downstream data as it is present only in pretraining but not in fine-tuning. In this work, we offer a new perspective on the consequence of such a discrepancy: We demonstrate empirically and theoretically that MLM pretraining allocates some model dimensions exclusively for representing [MASK] tokens, resulting in a representation deficiency for real tokens and limiting the pretrained model's expressiveness when it is adapted to downstream data without [MASK] tokens. Motivated by the identified issue, we propose MAE-LM, which pretrains the Masked Autoencoder architecture with MLM where [MASK] tokens are excluded from the encoder. Empirically, we show that MAE-LM improves the utilization of model dimensions for real token representations, and MAE-LM consistently outperforms MLM-pretrained models across different pretraining settings and model sizes when fine-tuned on the GLUE and SQuAD benchmarks.

This paper integrates graph-to-sequence into an end-to-end text-to-speech framework for syntax-aware modelling with syntactic information of input text. Specifically, the input text is parsed by a dependency parsing module to form a syntactic graph. The syntactic graph is then encoded by a graph encoder to extract the syntactic hidden information, which is concatenated with phoneme embedding and input to the alignment and flow-based decoding modules to generate the raw audio waveform. The model is experimented on two languages, English and Mandarin, using single-speaker, few samples of target speakers, and multi-speaker datasets, respectively. Experimental results show better prosodic consistency performance between input text and generated audio, and also get higher scores in the subjective prosodic evaluation, and show the ability of voice conversion. Besides, the efficiency of the model is largely boosted through the design of the AI chip operator with 5x acceleration.

Significant improvement has been achieved in automated audio captioning (AAC) with recent models. However, these models have become increasingly large as their performance is enhanced. In this work, we propose a knowledge distillation (KD) framework for AAC. Our analysis shows that in the encoder-decoder based AAC models, it is more effective to distill knowledge into the encoder as compared with the decoder. To this end, we incorporate encoder-level KD loss into training, in addition to the standard supervised loss and sequence-level KD loss. We investigate two encoder-level KD methods, based on mean squared error (MSE) loss and contrastive loss, respectively. Experimental results demonstrate that contrastive KD is more robust than MSE KD, exhibiting superior performance in data-scarce situations. By leveraging audio-only data into training in the KD framework, our student model achieves competitive performance, with an inference speed that is 19 times fasterAn online demo is available at \url{https://huggingface.co/spaces/wsntxxn/efficient_audio_captioning}.

The recent emergence of joint CTC-Attention model shows significant improvement in automatic speech recognition (ASR). The improvement largely lies in the modeling of linguistic information by decoder. The decoder joint-optimized with an acoustic encoder renders the language model from ground-truth sequences in an auto-regressive manner during training. However, the training corpus of the decoder is limited to the speech transcriptions, which is far less than the corpus needed to train an acceptable language model. This leads to poor robustness of decoder. To alleviate this problem, we propose linguistic-enhanced transformer, which introduces refined CTC information to decoder during training process, so that the decoder can be more robust. Our experiments on AISHELL-1 speech corpus show that the character error rate (CER) is relatively reduced by up to 7%. We also find that in joint CTC-Attention ASR model, decoder is more sensitive to linguistic information than acoustic information.

Video and audio are closely correlated modalities that humans naturally perceive together. While recent advancements have enabled the generation of audio or video from text, producing both modalities simultaneously still typically relies on either a cascaded process or multi-modal contrastive encoders. These approaches, however, often lead to suboptimal results due to inherent information losses during inference and conditioning. In this paper, we introduce SyncFlow, a system that is capable of simultaneously generating temporally synchronized audio and video from text. The core of SyncFlow is the proposed dual-diffusion-transformer (d-DiT) architecture, which enables joint video and audio modelling with proper information fusion. To efficiently manage the computational cost of joint audio and video modelling, SyncFlow utilizes a multi-stage training strategy that separates video and audio learning before joint fine-tuning. Our empirical evaluations demonstrate that SyncFlow produces audio and video outputs that are more correlated than baseline methods with significantly enhanced audio quality and audio-visual correspondence. Moreover, we demonstrate strong zero-shot capabilities of SyncFlow, including zero-shot video-to-audio generation and adaptation to novel video resolutions without further training.

We present Video-LLaMA, a multi-modal framework that empowers Large Language Models (LLMs) with the capability of understanding both visual and auditory content in the video. Video-LLaMA bootstraps cross-modal training from the frozen pre-trained visual \& audio encoders and the frozen LLMs. Unlike previous vision- LLMs that focus on static image comprehensions such as MiniGPT-4~zhu2023minigpt and LLaVA~liu2023visualit, Video-LLaMA tackles two challenges in video understanding: (1) capturing the temporal changes in visual scenes, (2) integrating audio-visual signals. For the first challenge, we propose Video Q-former to extend the pre-trained image encoder to a video encoder and introduce a video-to-text generation task to learn video-language correspondence. For the second challenge, we leverage ImageBind~girdhar2023imagebind as the pre-trained audio encoder which performs exceptionally well in aligning different modalities to a common embedding space. And then introduce an Audio Q-former to learn auditory query tokens. To align the output of both visual \& audio encoder with LLM's embedding space, we train Video-LLaMA on a large-scale vision caption dataset and a hign-quantity vision-instruction-tuning dataset. We found Video-LLaMA showcases the ability to perceive and comprehend video content, generating meaningful responses that are grounded in the visual and auditory information present in the videos. This highlights the potential of Video-LLaMA as a promising prototype for audio-visual AI assistants. Our code, pre-trained model, and demo are available at https://github.com/DAMO-NLP-SG/Video-LLaMA.

Recently, neural networks based purely on self-attention, such as the Vision Transformer (ViT), have been shown to outperform deep learning models constructed with convolutional neural networks (CNNs) on various vision tasks, thus extending the success of Transformers, which were originally developed for language processing, to the vision domain. A recent study showed that a similar methodology can also be applied to the audio domain. Specifically, the Audio Spectrogram Transformer (AST) achieves state-of-the-art results on various audio classification benchmarks. However, pure Transformer models tend to require more training data compared to CNNs, and the success of the AST relies on supervised pretraining that requires a large amount of labeled data and a complex training pipeline, thus limiting the practical usage of AST. This paper focuses on audio and speech classification, and aims to reduce the need for large amounts of labeled data for AST by leveraging self-supervised learning using unlabeled data. Specifically, we propose to pretrain the AST model with joint discriminative and generative masked spectrogram patch modeling (MSPM) using unlabeled audio from AudioSet and Librispeech. We evaluate our pretrained models on both audio and speech classification tasks including audio event classification, keyword spotting, emotion recognition, and speaker identification. The proposed self-supervised framework significantly boosts AST performance on all tasks, with an average improvement of 60.9%, leading to similar or even better results than a supervised pretrained AST. To the best of our knowledge, it is the first patch-based self-supervised learning framework in the audio and speech domain, and also the first self-supervised learning framework for AST.

This paper introduces a novel voice conversion (VC) model, guided by text instructions such as "articulate slowly with a deep tone" or "speak in a cheerful boyish voice". Unlike traditional methods that rely on reference utterances to determine the attributes of the converted speech, our model adds versatility and specificity to voice conversion. The proposed VC model is a neural codec language model which processes a sequence of discrete codes, resulting in the code sequence of converted speech. It utilizes text instructions as style prompts to modify the prosody and emotional information of the given speech. In contrast to previous approaches, which often rely on employing separate encoders like prosody and content encoders to handle different aspects of the source speech, our model handles various information of speech in an end-to-end manner. Experiments have demonstrated the impressive capabilities of our model in comprehending instructions and delivering reasonable results.

We introduce StyleFusion-TTS, a prompt and/or audio referenced, style and speaker-controllable, zero-shot text-to-speech (TTS) synthesis system designed to enhance the editability and naturalness of current research literature. We propose a general front-end encoder as a compact and effective module to utilize multimodal inputs including text prompts, audio references, and speaker timbre references in a fully zero-shot manner and produce disentangled style and speaker control embeddings. Our novel approach also leverages a hierarchical conformer structure for the fusion of style and speaker control embeddings, aiming to achieve optimal feature fusion within the current advanced TTS architecture. StyleFusion-TTS is evaluated through multiple metrics, both subjectively and objectively. The system shows promising performance across our evaluations, suggesting its potential to contribute to the advancement of the field of zero-shot text-to-speech synthesis.

Recent advances in neural text-to-speech research have been dominated by two-stage pipelines utilizing low-level intermediate speech representation such as mel-spectrograms. However, such predetermined features are fundamentally limited, because they do not allow to exploit the full potential of a data-driven approach through learning hidden representations. For this reason, several end-to-end methods have been proposed. However, such models are harder to train and require a large number of high-quality recordings with transcriptions. Here, we propose WavThruVec - a two-stage architecture that resolves the bottleneck by using high-dimensional Wav2Vec 2.0 embeddings as intermediate speech representation. Since these hidden activations provide high-level linguistic features, they are more robust to noise. That allows us to utilize annotated speech datasets of a lower quality to train the first-stage module. At the same time, the second-stage component can be trained on large-scale untranscribed audio corpora, as Wav2Vec 2.0 embeddings are already time-aligned. This results in an increased generalization capability to out-of-vocabulary words, as well as to a better generalization to unseen speakers. We show that the proposed model not only matches the quality of state-of-the-art neural models, but also presents useful properties enabling tasks like voice conversion or zero-shot synthesis.

Audio deepfake detection (ADD) is essential for preventing the misuse of synthetic voices that may infringe on personal rights and privacy. Recent zero-shot text-to-speech (TTS) models pose higher risks as they can clone voices with a single utterance. However, the existing ADD datasets are outdated, leading to suboptimal generalization of detection models. In this paper, we construct a new cross-domain ADD dataset comprising over 300 hours of speech data that is generated by five advanced zero-shot TTS models. To simulate real-world scenarios, we employ diverse attack methods and audio prompts from different datasets. Experiments show that, through novel attack-augmented training, the Wav2Vec2-large and Whisper-medium models achieve equal error rates of 4.1\% and 6.5\% respectively. Additionally, we demonstrate our models' outstanding few-shot ADD ability by fine-tuning with just one minute of target-domain data. Nonetheless, neural codec compressors greatly affect the detection accuracy, necessitating further research.

Audio-Visual Emotion Recognition (AVER) has garnered increasing attention in recent years for its critical role in creating emotion-ware intelligent machines. Previous efforts in this area are dominated by the supervised learning paradigm. Despite significant progress, supervised learning is meeting its bottleneck due to the longstanding data scarcity issue in AVER. Motivated by recent advances in self-supervised learning, we propose Hierarchical Contrastive Masked Autoencoder (HiCMAE), a novel self-supervised framework that leverages large-scale self-supervised pre-training on vast unlabeled audio-visual data to promote the advancement of AVER. Following prior arts in self-supervised audio-visual representation learning, HiCMAE adopts two primary forms of self-supervision for pre-training, namely masked data modeling and contrastive learning. Unlike them which focus exclusively on top-layer representations while neglecting explicit guidance of intermediate layers, HiCMAE develops a three-pronged strategy to foster hierarchical audio-visual feature learning and improve the overall quality of learned representations. To verify the effectiveness of HiCMAE, we conduct extensive experiments on 9 datasets covering both categorical and dimensional AVER tasks. Experimental results show that our method significantly outperforms state-of-the-art supervised and self-supervised audio-visual methods, which indicates that HiCMAE is a powerful audio-visual emotion representation learner. Codes and models will be publicly available at https://github.com/sunlicai/HiCMAE.

Motivated by the success of masked language modeling~(MLM) in pre-training natural language processing models, we propose w2v-BERT that explores MLM for self-supervised speech representation learning. w2v-BERT is a framework that combines contrastive learning and MLM, where the former trains the model to discretize input continuous speech signals into a finite set of discriminative speech tokens, and the latter trains the model to learn contextualized speech representations via solving a masked prediction task consuming the discretized tokens. In contrast to existing MLM-based speech pre-training frameworks such as HuBERT, which relies on an iterative re-clustering and re-training process, or vq-wav2vec, which concatenates two separately trained modules, w2v-BERT can be optimized in an end-to-end fashion by solving the two self-supervised tasks~(the contrastive task and MLM) simultaneously. Our experiments show that w2v-BERT achieves competitive results compared to current state-of-the-art pre-trained models on the LibriSpeech benchmarks when using the Libri-Light~60k corpus as the unsupervised data. In particular, when compared to published models such as conformer-based wav2vec~2.0 and HuBERT, our model shows~5\% to~10\% relative WER reduction on the test-clean and test-other subsets. When applied to the Google's Voice Search traffic dataset, w2v-BERT outperforms our internal conformer-based wav2vec~2.0 by more than~30\% relatively.

With the proliferation of Audio Language Model (ALM) based deepfake audio, there is an urgent need for effective detection methods. Unlike traditional deepfake audio generation, which often involves multi-step processes culminating in vocoder usage, ALM directly utilizes neural codec methods to decode discrete codes into audio. Moreover, driven by large-scale data, ALMs exhibit remarkable robustness and versatility, posing a significant challenge to current audio deepfake detection (ADD) models. To effectively detect ALM-based deepfake audio, we focus on the mechanism of the ALM-based audio generation method, the conversion from neural codec to waveform. We initially construct the Codecfake dataset, an open-source large-scale dataset, including two languages, millions of audio samples, and various test conditions, tailored for ALM-based audio detection. Additionally, to achieve universal detection of deepfake audio and tackle domain ascent bias issue of original SAM, we propose the CSAM strategy to learn a domain balanced and generalized minima. Experiment results demonstrate that co-training on Codecfake dataset and vocoded dataset with CSAM strategy yield the lowest average Equal Error Rate (EER) of 0.616% across all test conditions compared to baseline models.

We introduce SONAR, a new multilingual and multimodal fixed-size sentence embedding space. Our single text encoder, covering 200 languages, substantially outperforms existing sentence embeddings such as LASER3 and LabSE on the xsim and xsim++ multilingual similarity search tasks. Speech segments can be embedded in the same SONAR embedding space using language-specific speech encoders trained in a teacher-student setting on speech transcription data. Our encoders outperform existing speech encoders on similarity search tasks. We also provide a text decoder for 200 languages, which allows us to perform text-to-text and speech-to-text machine translation, including for zero-shot language and modality combinations. Our text-to-text results are competitive compared to the state-of-the-art NLLB~1B model, despite the fixed-size bottleneck representation. Our zero-shot speech-to-text translation results compare favorably with strong supervised baselines such as Whisper.

This paper proposes a simple yet effective way of regularising the encoder-decoder-based automatic speech recognition (ASR) models that enhance the robustness of the model and improve the generalisation to out-of-domain scenarios. The proposed approach is dubbed as Decoder-Centric Regularisation in Encoder-Decoder (DeCRED) architecture for ASR, where auxiliary classifier(s) is introduced in layers of the decoder module. Leveraging these classifiers, we propose two decoding strategies that re-estimate the next token probabilities. Using the recent E-branchformer architecture, we build strong ASR systems that obtained competitive WERs as compared to Whisper-medium and outperformed OWSM v3; while relying only on a fraction of training data and model size. On top of such a strong baseline, we show that DeCRED can further improve the results and, moreover, generalise much better to out-of-domain scenarios, where we show an absolute reduction of 2.7 and 2.9 WERs on AMI and Gigaspeech datasets, respectively. We provide extensive analysis and accompanying experiments that support the benefits of the proposed regularisation scheme.

We introduce VoiceCraft, a token infilling neural codec language model, that achieves state-of-the-art performance on both speech editing and zero-shot text-to-speech (TTS) on audiobooks, internet videos, and podcasts. VoiceCraft employs a Transformer decoder architecture and introduces a token rearrangement procedure that combines causal masking and delayed stacking to enable generation within an existing sequence. On speech editing tasks, VoiceCraft produces edited speech that is nearly indistinguishable from unedited recordings in terms of naturalness, as evaluated by humans; for zero-shot TTS, our model outperforms prior SotA models including VALLE and the popular commercial model XTTS-v2. Crucially, the models are evaluated on challenging and realistic datasets, that consist of diverse accents, speaking styles, recording conditions, and background noise and music, and our model performs consistently well compared to other models and real recordings. In particular, for speech editing evaluation, we introduce a high quality, challenging, and realistic dataset named RealEdit. We encourage readers to listen to the demos at https://jasonppy.github.io/VoiceCraft_web.

Recent works have shown promising results in connecting speech encoders to large language models (LLMs) for speech recognition. However, several limitations persist, including limited fine-tuning options, a lack of mechanisms to enforce speech-text alignment, and high insertion errors especially in domain mismatch conditions. This paper presents a comprehensive solution to address these issues. We begin by investigating more thoughtful fine-tuning schemes. Next, we propose a matching loss to enhance alignment between modalities. Finally, we explore training and inference methods to mitigate high insertion errors. Experimental results on the Librispeech corpus demonstrate that partially fine-tuning the encoder and LLM using parameter-efficient methods, such as LoRA, is the most cost-effective approach. Additionally, the matching loss improves modality alignment, enhancing performance. The proposed training and inference methods significantly reduce insertion errors.

Audio language models have recently emerged as a promising approach for various audio generation tasks, relying on audio tokenizers to encode waveforms into sequences of discrete symbols. Audio tokenization often poses a necessary compromise between code bitrate and reconstruction accuracy. When dealing with low-bitrate audio codes, language models are constrained to process only a subset of the information embedded in the audio, which in turn restricts their generative capabilities. To circumvent these issues, we propose encoding audio as vector sequences in continuous space mathbb R^d and autoregressively generating these sequences using a decoder-only diffusion transformer (ARDiT). Our findings indicate that ARDiT excels in zero-shot text-to-speech and exhibits performance that compares to or even surpasses that of state-of-the-art models. High-bitrate continuous speech representation enables almost flawless reconstruction, allowing our model to achieve nearly perfect speech editing. Our experiments reveal that employing Integral Kullback-Leibler (IKL) divergence for distillation at each autoregressive step significantly boosts the perceived quality of the samples. Simultaneously, it condenses the iterative sampling process of the diffusion model into a single step. Furthermore, ARDiT can be trained to predict several continuous vectors in one step, significantly reducing latency during sampling. Impressively, one of our models can generate 170 ms of 24 kHz speech per evaluation step with minimal degradation in performance. Audio samples are available at http://ardit-tts.github.io/ .

We present a causal speech enhancement model working on the raw waveform that runs in real-time on a laptop CPU. The proposed model is based on an encoder-decoder architecture with skip-connections. It is optimized on both time and frequency domains, using multiple loss functions. Empirical evidence shows that it is capable of removing various kinds of background noise including stationary and non-stationary noises, as well as room reverb. Additionally, we suggest a set of data augmentation techniques applied directly on the raw waveform which further improve model performance and its generalization abilities. We perform evaluations on several standard benchmarks, both using objective metrics and human judgements. The proposed model matches state-of-the-art performance of both causal and non causal methods while working directly on the raw waveform.

We introduce EgoSonics, a method to generate semantically meaningful and synchronized audio tracks conditioned on silent egocentric videos. Generating audio for silent egocentric videos could open new applications in virtual reality, assistive technologies, or for augmenting existing datasets. Existing work has been limited to domains like speech, music, or impact sounds and cannot easily capture the broad range of audio frequencies found in egocentric videos. EgoSonics addresses these limitations by building on the strength of latent diffusion models for conditioned audio synthesis. We first encode and process audio and video data into a form that is suitable for generation. The encoded data is used to train our model to generate audio tracks that capture the semantics of the input video. Our proposed SyncroNet builds on top of ControlNet to provide control signals that enables temporal synchronization to the synthesized audio. Extensive evaluations show that our model outperforms existing work in audio quality, and in our newly proposed synchronization evaluation method. Furthermore, we demonstrate downstream applications of our model in improving video summarization.

Automated Audio Captioning (AAC) aims to develop systems capable of describing an audio recording using a textual sentence. In contrast, Audio-Text Retrieval (ATR) systems seek to find the best matching audio recording(s) for a given textual query (Text-to-Audio) or vice versa (Audio-to-Text). These tasks require different types of systems: AAC employs a sequence-to-sequence model, while ATR utilizes a ranking model that compares audio and text representations within a shared projection subspace. However, this work investigates the relationship between AAC and ATR by exploring the ATR capabilities of an unmodified AAC system, without fine-tuning for the new task. Our AAC system consists of an audio encoder (ConvNeXt-Tiny) trained on AudioSet for audio tagging, and a transformer decoder responsible for generating sentences. For AAC, it achieves a high SPIDEr-FL score of 0.298 on Clotho and 0.472 on AudioCaps on average. For ATR, we propose using the standard Cross-Entropy loss values obtained for any audio/caption pair. Experimental results on the Clotho and AudioCaps datasets demonstrate decent recall values using this simple approach. For instance, we obtained a Text-to-Audio R@1 value of 0.382 for Au-dioCaps, which is above the current state-of-the-art method without external data. Interestingly, we observe that normalizing the loss values was necessary for Audio-to-Text retrieval.

We introduce CLaMP: Contrastive Language-Music Pre-training, which learns cross-modal representations between natural language and symbolic music using a music encoder and a text encoder trained jointly with a contrastive loss. To pre-train CLaMP, we collected a large dataset of 1.4 million music-text pairs. It employed text dropout as a data augmentation technique and bar patching to efficiently represent music data which reduces sequence length to less than 10%. In addition, we developed a masked music model pre-training objective to enhance the music encoder's comprehension of musical context and structure. CLaMP integrates textual information to enable semantic search and zero-shot classification for symbolic music, surpassing the capabilities of previous models. To support the evaluation of semantic search and music classification, we publicly release WikiMusicText (WikiMT), a dataset of 1010 lead sheets in ABC notation, each accompanied by a title, artist, genre, and description. In comparison to state-of-the-art models that require fine-tuning, zero-shot CLaMP demonstrated comparable or superior performance on score-oriented datasets.

In this work, we present and evaluate SELMA, a Speech-Enabled Language Model for virtual Assistant interactions that integrates audio and text as inputs to a Large Language Model (LLM). SELMA is designed to handle three primary and two auxiliary tasks related to interactions with virtual assistants simultaneously within a single end-to-end model. We employ low-rank adaptation modules for parameter-efficient training of both the audio encoder and the LLM. Additionally, we implement a feature pooling strategy enabling the system to recognize global patterns and improve accuracy on tasks less reliant on individual sequence elements. Experimental results on Voice Trigger (VT) detection, Device-Directed Speech Detection (DDSD), and Automatic Speech Recognition (ASR), demonstrate that our approach both simplifies the typical input processing pipeline of virtual assistants significantly and also improves performance compared to dedicated models for each individual task. SELMA yields relative Equal-Error Rate improvements of 64% on the VT detection task, and 22% on DDSD, while also achieving word error rates close to the baseline.

We present a fast and high-quality codec language model for parallel audio generation. While SoundStorm, a state-of-the-art parallel audio generation model, accelerates inference speed compared to autoregressive models, it still suffers from slow inference due to iterative sampling. To resolve this problem, we propose Group-Masked Language Modeling~(G-MLM) and Group Iterative Parallel Decoding~(G-IPD) for efficient parallel audio generation. Both the training and sampling schemes enable the model to synthesize high-quality audio with a small number of iterations by effectively modeling the group-wise conditional dependencies. In addition, our model employs a cross-attention-based architecture to capture the speaker style of the prompt voice and improves computational efficiency. Experimental results demonstrate that our proposed model outperforms the baselines in prompt-based audio generation.

Non-parallel many-to-many voice conversion remains an interesting but challenging speech processing task. Recently, AutoVC, a conditional autoencoder based method, achieved excellent conversion results by disentangling the speaker identity and the speech content using information-constraining bottlenecks. However, due to the pure autoencoder training method, it is difficult to evaluate the separation effect of content and speaker identity. In this paper, a novel voice conversion framework, named boldsymbol Text boldsymbol Guided boldsymbol AutoVC(TGAVC), is proposed to more effectively separate content and timbre from speech, where an expected content embedding produced based on the text transcriptions is designed to guide the extraction of voice content. In addition, the adversarial training is applied to eliminate the speaker identity information in the estimated content embedding extracted from speech. Under the guidance of the expected content embedding and the adversarial training, the content encoder is trained to extract speaker-independent content embedding from speech. Experiments on AIShell-3 dataset show that the proposed model outperforms AutoVC in terms of naturalness and similarity of converted speech.

The remarkable performance achieved by Large Language Models (LLM) has driven research efforts to leverage them for a wide range of tasks and input modalities. In speech-to-text (S2T) tasks, the emerging solution consists of projecting the output of the encoder of a Speech Foundational Model (SFM) into the LLM embedding space through an adapter module. However, no work has yet investigated how much the downstream-task performance depends on each component (SFM, adapter, LLM) nor whether the best design of the adapter depends on the chosen SFM and LLM. To fill this gap, we evaluate the combination of 5 adapter modules, 2 LLMs (Mistral and Llama), and 2 SFMs (Whisper and SeamlessM4T) on two widespread S2T tasks, namely Automatic Speech Recognition and Speech Translation. Our results demonstrate that the SFM plays a pivotal role in downstream performance, while the adapter choice has moderate impact and depends on the SFM and LLM.

In recent years, diffusion-based text-to-music (TTM) generation has gained prominence, offering an innovative approach to synthesizing musical content from textual descriptions. Achieving high accuracy and diversity in this generation process requires extensive, high-quality data, including both high-fidelity audio waveforms and detailed text descriptions, which often constitute only a small portion of available datasets. In open-source datasets, issues such as low-quality music waveforms, mislabeling, weak labeling, and unlabeled data significantly hinder the development of music generation models. To address these challenges, we propose a novel paradigm for high-quality music generation that incorporates a quality-aware training strategy, enabling generative models to discern the quality of input music waveforms during training. Leveraging the unique properties of musical signals, we first adapted and implemented a masked diffusion transformer (MDT) model for the TTM task, demonstrating its distinct capacity for quality control and enhanced musicality. Additionally, we address the issue of low-quality captions in TTM with a caption refinement data processing approach. Experiments demonstrate our state-of-the-art (SOTA) performance on MusicCaps and the Song-Describer Dataset. Our demo page can be accessed at https://qa-mdt.github.io/.

In this work, we present a hybrid CTC/Attention model based on a ResNet-18 and Convolution-augmented transformer (Conformer), that can be trained in an end-to-end manner. In particular, the audio and visual encoders learn to extract features directly from raw pixels and audio waveforms, respectively, which are then fed to conformers and then fusion takes place via a Multi-Layer Perceptron (MLP). The model learns to recognise characters using a combination of CTC and an attention mechanism. We show that end-to-end training, instead of using pre-computed visual features which is common in the literature, the use of a conformer, instead of a recurrent network, and the use of a transformer-based language model, significantly improve the performance of our model. We present results on the largest publicly available datasets for sentence-level speech recognition, Lip Reading Sentences 2 (LRS2) and Lip Reading Sentences 3 (LRS3), respectively. The results show that our proposed models raise the state-of-the-art performance by a large margin in audio-only, visual-only, and audio-visual experiments.

Given a script, the challenge in Movie Dubbing (Visual Voice Cloning, V2C) is to generate speech that aligns well with the video in both time and emotion, based on the tone of a reference audio track. Existing state-of-the-art V2C models break the phonemes in the script according to the divisions between video frames, which solves the temporal alignment problem but leads to incomplete phoneme pronunciation and poor identity stability. To address this problem, we propose StyleDubber, which switches dubbing learning from the frame level to phoneme level. It contains three main components: (1) A multimodal style adaptor operating at the phoneme level to learn pronunciation style from the reference audio, and generate intermediate representations informed by the facial emotion presented in the video; (2) An utterance-level style learning module, which guides both the mel-spectrogram decoding and the refining processes from the intermediate embeddings to improve the overall style expression; And (3) a phoneme-guided lip aligner to maintain lip sync. Extensive experiments on two of the primary benchmarks, V2C and Grid, demonstrate the favorable performance of the proposed method as compared to the current state-of-the-art. The source code and trained models will be released to the public.

Recent advances in text-to-speech have significantly improved the expressiveness of synthetic speech. However, a major challenge remains in generating speech that captures the diverse styles exhibited by professional narrators in audiobooks without relying on manually labeled data or reference speech.To address this problem, we propose a text-aware and context-aware(TACA) style modeling approach for expressive audiobook speech synthesis. We first establish a text-aware style space to cover diverse styles via contrastive learning with the supervision of the speech style. Meanwhile, we adopt a context encoder to incorporate cross-sentence information and the style embedding obtained from text. Finally, we introduce the context encoder to two typical TTS models, VITS-based TTS and language model-based TTS. Experimental results demonstrate that our proposed approach can effectively capture diverse styles and coherent prosody, and consequently improves naturalness and expressiveness in audiobook speech synthesis.

We consider the task of generating diverse and realistic videos guided by natural audio samples from a wide variety of semantic classes. For this task, the videos are required to be aligned both globally and temporally with the input audio: globally, the input audio is semantically associated with the entire output video, and temporally, each segment of the input audio is associated with a corresponding segment of that video. We utilize an existing text-conditioned video generation model and a pre-trained audio encoder model. The proposed method is based on a lightweight adaptor network, which learns to map the audio-based representation to the input representation expected by the text-to-video generation model. As such, it also enables video generation conditioned on text, audio, and, for the first time as far as we can ascertain, on both text and audio. We validate our method extensively on three datasets demonstrating significant semantic diversity of audio-video samples and further propose a novel evaluation metric (AV-Align) to assess the alignment of generated videos with input audio samples. AV-Align is based on the detection and comparison of energy peaks in both modalities. In comparison to recent state-of-the-art approaches, our method generates videos that are better aligned with the input sound, both with respect to content and temporal axis. We also show that videos produced by our method present higher visual quality and are more diverse.

Interactions with virtual assistants typically start with a trigger phrase followed by a command. In this work, we explore the possibility of making these interactions more natural by eliminating the need for a trigger phrase. Our goal is to determine whether a user addressed the virtual assistant based on signals obtained from the streaming audio recorded by the device microphone. We address this task by combining 1-best hypotheses and decoder signals from an automatic speech recognition system with acoustic representations from an audio encoder as input features to a large language model (LLM). In particular, we are interested in data and resource efficient systems that require only a small amount of training data and can operate in scenarios with only a single frozen LLM available on a device. For this reason, our model is trained on 80k or less examples of multimodal data using a combination of low-rank adaptation and prefix tuning. We compare the proposed system to unimodal baselines and show that the multimodal approach achieves lower equal-error-rates (EERs), while using only a fraction of the training data. We also show that low-dimensional specialized audio representations lead to lower EERs than high-dimensional general audio representations.

This survey paper provides a comprehensive overview of the recent advancements and challenges in applying large language models to the field of audio signal processing. Audio processing, with its diverse signal representations and a wide range of sources--from human voices to musical instruments and environmental sounds--poses challenges distinct from those found in traditional Natural Language Processing scenarios. Nevertheless, Large Audio Models, epitomized by transformer-based architectures, have shown marked efficacy in this sphere. By leveraging massive amount of data, these models have demonstrated prowess in a variety of audio tasks, spanning from Automatic Speech Recognition and Text-To-Speech to Music Generation, among others. Notably, recently these Foundational Audio Models, like SeamlessM4T, have started showing abilities to act as universal translators, supporting multiple speech tasks for up to 100 languages without any reliance on separate task-specific systems. This paper presents an in-depth analysis of state-of-the-art methodologies regarding Foundational Large Audio Models, their performance benchmarks, and their applicability to real-world scenarios. We also highlight current limitations and provide insights into potential future research directions in the realm of Large Audio Models with the intent to spark further discussion, thereby fostering innovation in the next generation of audio-processing systems. Furthermore, to cope with the rapid development in this area, we will consistently update the relevant repository with relevant recent articles and their open-source implementations at https://github.com/EmulationAI/awesome-large-audio-models.

Large-scale pre-trained language models have been shown to be helpful in improving the naturalness of text-to-speech (TTS) models by enabling them to produce more naturalistic prosodic patterns. However, these models are usually word-level or sup-phoneme-level and jointly trained with phonemes, making them inefficient for the downstream TTS task where only phonemes are needed. In this work, we propose a phoneme-level BERT (PL-BERT) with a pretext task of predicting the corresponding graphemes along with the regular masked phoneme predictions. Subjective evaluations show that our phoneme-level BERT encoder has significantly improved the mean opinion scores (MOS) of rated naturalness of synthesized speech compared with the state-of-the-art (SOTA) StyleTTS baseline on out-of-distribution (OOD) texts.

Recent advancements in diffusion models and large language models (LLMs) have significantly propelled the field of AIGC. Text-to-Audio (TTA), a burgeoning AIGC application designed to generate audio from natural language prompts, is attracting increasing attention. However, existing TTA studies often struggle with generation quality and text-audio alignment, especially for complex textual inputs. Drawing inspiration from state-of-the-art Text-to-Image (T2I) diffusion models, we introduce Auffusion, a TTA system adapting T2I model frameworks to TTA task, by effectively leveraging their inherent generative strengths and precise cross-modal alignment. Our objective and subjective evaluations demonstrate that Auffusion surpasses previous TTA approaches using limited data and computational resource. Furthermore, previous studies in T2I recognizes the significant impact of encoder choice on cross-modal alignment, like fine-grained details and object bindings, while similar evaluation is lacking in prior TTA works. Through comprehensive ablation studies and innovative cross-attention map visualizations, we provide insightful assessments of text-audio alignment in TTA. Our findings reveal Auffusion's superior capability in generating audios that accurately match textual descriptions, which further demonstrated in several related tasks, such as audio style transfer, inpainting and other manipulations. Our implementation and demos are available at https://auffusion.github.io.

Diffusion models have emerged as the new state-of-the-art family of deep generative models, and their promising potentials for text generation have recently attracted increasing attention. Existing studies mostly adopt a single encoder architecture with partially noising processes for conditional text generation, but its degree of flexibility for conditional modeling is limited. In fact, the encoder-decoder architecture is naturally more flexible for its detachable encoder and decoder modules, which is extensible to multilingual and multimodal generation tasks for conditions and target texts. However, the encoding process of conditional texts lacks the understanding of target texts. To this end, a spiral interaction architecture for encoder-decoder text diffusion (DiffuSIA) is proposed. Concretely, the conditional information from encoder is designed to be captured by the diffusion decoder, while the target information from decoder is designed to be captured by the conditional encoder. These two types of information flow run through multilayer interaction spirally for deep fusion and understanding. DiffuSIA is evaluated on four text generation tasks, including paraphrase, text simplification, question generation, and open-domain dialogue generation. Experimental results show that DiffuSIA achieves competitive performance among previous methods on all four tasks, demonstrating the effectiveness and generalization ability of the proposed method.

Masked language modeling is widely used for pretraining large language models for natural language understanding (NLU). However, random masking is suboptimal, allocating an equal masking rate for all tokens. In this paper, we propose InforMask, a new unsupervised masking strategy for training masked language models. InforMask exploits Pointwise Mutual Information (PMI) to select the most informative tokens to mask. We further propose two optimizations for InforMask to improve its efficiency. With a one-off preprocessing step, InforMask outperforms random masking and previously proposed masking strategies on the factual recall benchmark LAMA and the question answering benchmark SQuAD v1 and v2.

The massive growth of self-supervised learning (SSL) has been witnessed in language, vision, speech, and audio domains over the past few years. While discrete label prediction is widely adopted for other modalities, the state-of-the-art audio SSL models still employ reconstruction loss for pre-training. Compared with reconstruction loss, semantic-rich discrete label prediction encourages the SSL model to abstract the high-level audio semantics and discard the redundant details as in human perception. However, a semantic-rich acoustic tokenizer for general audio pre-training is usually not straightforward to obtain, due to the continuous property of audio and unavailable phoneme sequences like speech. To tackle this challenge, we propose BEATs, an iterative audio pre-training framework to learn Bidirectional Encoder representation from Audio Transformers, where an acoustic tokenizer and an audio SSL model are optimized by iterations. In the first iteration, we use random projection as the acoustic tokenizer to train an audio SSL model in a mask and label prediction manner. Then, we train an acoustic tokenizer for the next iteration by distilling the semantic knowledge from the pre-trained or fine-tuned audio SSL model. The iteration is repeated with the hope of mutual promotion of the acoustic tokenizer and audio SSL model. The experimental results demonstrate our acoustic tokenizers can generate discrete labels with rich audio semantics and our audio SSL models achieve state-of-the-art results across various audio classification benchmarks, even outperforming previous models that use more training data and model parameters significantly. Specifically, we set a new state-of-the-art mAP 50.6% on AudioSet-2M for audio-only models without using any external data, and 98.1% accuracy on ESC-50. The code and pre-trained models are available at https://aka.ms/beats.

Domain adaptation using text-only corpus is challenging in end-to-end(E2E) speech recognition. Adaptation by synthesizing audio from text through TTS is resource-consuming. We present a method to learn Unified Speech-Text Representation in Conformer Transducer(USTR-CT) to enable fast domain adaptation using the text-only corpus. Different from the previous textogram method, an extra text encoder is introduced in our work to learn text representation and is removed during inference, so there is no modification for online deployment. To improve the efficiency of adaptation, single-step and multi-step adaptations are also explored. The experiments on adapting LibriSpeech to SPGISpeech show the proposed method reduces the word error rate(WER) by relatively 44% on the target domain, which is better than those of TTS method and textogram method. Also, it is shown the proposed method can be combined with internal language model estimation(ILME) to further improve the performance.

We study Neural Foley, the automatic generation of high-quality sound effects synchronizing with videos, enabling an immersive audio-visual experience. Despite its wide range of applications, existing approaches encounter limitations when it comes to simultaneously synthesizing high-quality and video-aligned (i.e.,, semantic relevant and temporal synchronized) sounds. To overcome these limitations, we propose FoleyCrafter, a novel framework that leverages a pre-trained text-to-audio model to ensure high-quality audio generation. FoleyCrafter comprises two key components: the semantic adapter for semantic alignment and the temporal controller for precise audio-video synchronization. The semantic adapter utilizes parallel cross-attention layers to condition audio generation on video features, producing realistic sound effects that are semantically relevant to the visual content. Meanwhile, the temporal controller incorporates an onset detector and a timestampbased adapter to achieve precise audio-video alignment. One notable advantage of FoleyCrafter is its compatibility with text prompts, enabling the use of text descriptions to achieve controllable and diverse video-to-audio generation according to user intents. We conduct extensive quantitative and qualitative experiments on standard benchmarks to verify the effectiveness of FoleyCrafter. Models and codes are available at https://github.com/open-mmlab/FoleyCrafter.

We present AdVerb, a novel audio-visual dereverberation framework that uses visual cues in addition to the reverberant sound to estimate clean audio. Although audio-only dereverberation is a well-studied problem, our approach incorporates the complementary visual modality to perform audio dereverberation. Given an image of the environment where the reverberated sound signal has been recorded, AdVerb employs a novel geometry-aware cross-modal transformer architecture that captures scene geometry and audio-visual cross-modal relationship to generate a complex ideal ratio mask, which, when applied to the reverberant audio predicts the clean sound. The effectiveness of our method is demonstrated through extensive quantitative and qualitative evaluations. Our approach significantly outperforms traditional audio-only and audio-visual baselines on three downstream tasks: speech enhancement, speech recognition, and speaker verification, with relative improvements in the range of 18% - 82% on the LibriSpeech test-clean set. We also achieve highly satisfactory RT60 error scores on the AVSpeech dataset.

We show for the first time that learning powerful representations from speech audio alone followed by fine-tuning on transcribed speech can outperform the best semi-supervised methods while being conceptually simpler. wav2vec 2.0 masks the speech input in the latent space and solves a contrastive task defined over a quantization of the latent representations which are jointly learned. Experiments using all labeled data of Librispeech achieve 1.8/3.3 WER on the clean/other test sets. When lowering the amount of labeled data to one hour, wav2vec 2.0 outperforms the previous state of the art on the 100 hour subset while using 100 times less labeled data. Using just ten minutes of labeled data and pre-training on 53k hours of unlabeled data still achieves 4.8/8.2 WER. This demonstrates the feasibility of speech recognition with limited amounts of labeled data.

Language models (LMs) have recently flourished in natural language processing and computer vision, generating high-fidelity texts or images in various tasks. In contrast, the current speech generative models are still struggling regarding speech quality and task generalization. This paper presents Vec-Tok Speech, an extensible framework that resembles multiple speech generation tasks, generating expressive and high-fidelity speech. Specifically, we propose a novel speech codec based on speech vectors and semantic tokens. Speech vectors contain acoustic details contributing to high-fidelity speech reconstruction, while semantic tokens focus on the linguistic content of speech, facilitating language modeling. Based on the proposed speech codec, Vec-Tok Speech leverages an LM to undertake the core of speech generation. Moreover, Byte-Pair Encoding (BPE) is introduced to reduce the token length and bit rate for lower exposure bias and longer context coverage, improving the performance of LMs. Vec-Tok Speech can be used for intra- and cross-lingual zero-shot voice conversion (VC), zero-shot speaking style transfer text-to-speech (TTS), speech-to-speech translation (S2ST), speech denoising, and speaker de-identification and anonymization. Experiments show that Vec-Tok Speech, built on 50k hours of speech, performs better than other SOTA models. Code will be available at https://github.com/BakerBunker/VecTok .

In this work, we build a simple but strong baseline for sounding video generation. Given base diffusion models for audio and video, we integrate them with additional modules into a single model and train it to make the model jointly generate audio and video. To enhance alignment between audio-video pairs, we introduce two novel mechanisms in our model. The first one is timestep adjustment, which provides different timestep information to each base model. It is designed to align how samples are generated along with timesteps across modalities. The second one is a new design of the additional modules, termed Cross-Modal Conditioning as Positional Encoding (CMC-PE). In CMC-PE, cross-modal information is embedded as if it represents temporal position information, and the embeddings are fed into the model like positional encoding. Compared with the popular cross-attention mechanism, CMC-PE provides a better inductive bias for temporal alignment in the generated data. Experimental results validate the effectiveness of the two newly introduced mechanisms and also demonstrate that our method outperforms existing methods.

Music tagging is a task to predict the tags of music recordings. However, previous music tagging research primarily focuses on close-set music tagging tasks which can not be generalized to new tags. In this work, we propose a zero-shot music tagging system modeled by a joint music and language attention (JMLA) model to address the open-set music tagging problem. The JMLA model consists of an audio encoder modeled by a pretrained masked autoencoder and a decoder modeled by a Falcon7B. We introduce preceiver resampler to convert arbitrary length audio into fixed length embeddings. We introduce dense attention connections between encoder and decoder layers to improve the information flow between the encoder and decoder layers. We collect a large-scale music and description dataset from the internet. We propose to use ChatGPT to convert the raw descriptions into formalized and diverse descriptions to train the JMLA models. Our proposed JMLA system achieves a zero-shot audio tagging accuracy of 64.82% on the GTZAN dataset, outperforming previous zero-shot systems and achieves comparable results to previous systems on the FMA and the MagnaTagATune datasets.

Deep neural networks have become an indispensable technique for audio source separation (ASS). It was recently reported that a variant of CNN architecture called MMDenseNet was successfully employed to solve the ASS problem of estimating source amplitudes, and state-of-the-art results were obtained for DSD100 dataset. To further enhance MMDenseNet, here we propose a novel architecture that integrates long short-term memory (LSTM) in multiple scales with skip connections to efficiently model long-term structures within an audio context. The experimental results show that the proposed method outperforms MMDenseNet, LSTM and a blend of the two networks. The number of parameters and processing time of the proposed model are significantly less than those for simple blending. Furthermore, the proposed method yields better results than those obtained using ideal binary masks for a singing voice separation task.

Current state-of-the-art (SOTA) codec-based audio synthesis systems can mimic anyone's voice with just a 3-second sample from that specific unseen speaker. Unfortunately, malicious attackers may exploit these technologies, causing misuse and security issues. Anti-spoofing models have been developed to detect fake speech. However, the open question of whether current SOTA anti-spoofing models can effectively counter deepfake audios from codec-based speech synthesis systems remains unanswered. In this paper, we curate an extensive collection of contemporary SOTA codec models, employing them to re-create synthesized speech. This endeavor leads to the creation of CodecFake, the first codec-based deepfake audio dataset. Additionally, we verify that anti-spoofing models trained on commonly used datasets cannot detect synthesized speech from current codec-based speech generation systems. The proposed CodecFake dataset empowers these models to counter this challenge effectively.

Transformers have gained increasing popularity in a wide range of applications, including Natural Language Processing (NLP), Computer Vision and Speech Recognition, because of their powerful representational capacity. However, harnessing this representational capacity effectively requires a large amount of data, strong regularization, or both, to mitigate overfitting. Recently, the power of the Transformer has been unlocked by self-supervised pretraining strategies based on masked autoencoders which rely on reconstructing masked inputs, directly, or contrastively from unmasked content. This pretraining strategy which has been used in BERT models in NLP, Wav2Vec models in Speech and, recently, in MAE models in Vision, forces the model to learn about relationships between the content in different parts of the input using autoencoding related objectives. In this paper, we propose a novel, but surprisingly simple alternative to content reconstruction~-- that of predicting locations from content, without providing positional information for it. Doing so requires the Transformer to understand the positional relationships between different parts of the input, from their content alone. This amounts to an efficient implementation where the pretext task is a classification problem among all possible positions for each input token. We experiment on both Vision and Speech benchmarks, where our approach brings improvements over strong supervised training baselines and is comparable to modern unsupervised/self-supervised pretraining methods. Our method also enables Transformers trained without position embeddings to outperform ones trained with full position information.

Generating high-quality and person-generic visual dubbing remains a challenge. Recent innovation has seen the advent of a two-stage paradigm, decoupling the rendering and lip synchronization process facilitated by intermediate representation as a conduit. Still, previous methodologies rely on rough landmarks or are confined to a single speaker, thus limiting their performance. In this paper, we propose DiffDub: Diffusion-based dubbing. We first craft the Diffusion auto-encoder by an inpainting renderer incorporating a mask to delineate editable zones and unaltered regions. This allows for seamless filling of the lower-face region while preserving the remaining parts. Throughout our experiments, we encountered several challenges. Primarily, the semantic encoder lacks robustness, constricting its ability to capture high-level features. Besides, the modeling ignored facial positioning, causing mouth or nose jitters across frames. To tackle these issues, we employ versatile strategies, including data augmentation and supplementary eye guidance. Moreover, we encapsulated a conformer-based reference encoder and motion generator fortified by a cross-attention mechanism. This enables our model to learn person-specific textures with varying references and reduces reliance on paired audio-visual data. Our rigorous experiments comprehensively highlight that our ground-breaking approach outpaces existing methods with considerable margins and delivers seamless, intelligible videos in person-generic and multilingual scenarios.

This paper proposes a generative pretraining foundation model for high-quality speech restoration tasks. By directly operating on complex-valued short-time Fourier transform coefficients, our model does not rely on any vocoders for time-domain signal reconstruction. As a result, our model simplifies the synthesis process and removes the quality upper-bound introduced by any mel-spectrogram vocoder compared to prior work SpeechFlow. The proposed method is evaluated on multiple speech restoration tasks, including speech denoising, bandwidth extension, codec artifact removal, and target speaker extraction. In all scenarios, finetuning our pretrained model results in superior performance over strong baselines. Notably, in the target speaker extraction task, our model outperforms existing systems, including those leveraging SSL-pretrained encoders like WavLM. The code and the pretrained checkpoints are publicly available in the NVIDIA NeMo framework.

Despite recent progress in text-to-audio (TTA) generation, we show that the state-of-the-art models, such as AudioLDM, trained on datasets with an imbalanced class distribution, such as AudioCaps, are biased in their generation performance. Specifically, they excel in generating common audio classes while underperforming in the rare ones, thus degrading the overall generation performance. We refer to this problem as long-tailed text-to-audio generation. To address this issue, we propose a simple retrieval-augmented approach for TTA models. Specifically, given an input text prompt, we first leverage a Contrastive Language Audio Pretraining (CLAP) model to retrieve relevant text-audio pairs. The features of the retrieved audio-text data are then used as additional conditions to guide the learning of TTA models. We enhance AudioLDM with our proposed approach and denote the resulting augmented system as Re-AudioLDM. On the AudioCaps dataset, Re-AudioLDM achieves a state-of-the-art Frechet Audio Distance (FAD) of 1.37, outperforming the existing approaches by a large margin. Furthermore, we show that Re-AudioLDM can generate realistic audio for complex scenes, rare audio classes, and even unseen audio types, indicating its potential in TTA tasks.

Large language models (LLMs) have achieved remarkable success in the field of natural language processing, enabling better human-computer interaction using natural language. However, the seamless integration of speech signals into LLMs has not been explored well. The "decoder-only" architecture has also not been well studied for speech processing tasks. In this research, we introduce Speech-LLaMA, a novel approach that effectively incorporates acoustic information into text-based large language models. Our method leverages Connectionist Temporal Classification and a simple audio encoder to map the compressed acoustic features to the continuous semantic space of the LLM. In addition, we further probe the decoder-only architecture for speech-to-text tasks by training a smaller scale randomly initialized speech-LLaMA model from speech-text paired data alone. We conduct experiments on multilingual speech-to-text translation tasks and demonstrate a significant improvement over strong baselines, highlighting the potential advantages of decoder-only models for speech-to-text conversion.

Large Language Models (LLMs) demonstrate impressive capabilities, yet interaction with these models is mostly facilitated through text. Using Text-To-Speech to synthesize LLM outputs typically results in notable latency, which is impractical for fluent voice conversations. We propose LLM2Speech, an architecture to synthesize speech while text is being generated by an LLM which yields significant latency reduction. LLM2Speech mimics the predictions of a non-streaming teacher model while limiting the exposure to future context in order to enable streaming. It exploits the hidden embeddings of the LLM, a by-product of the text generation that contains informative semantic context. Experimental results show that LLM2Speech maintains the teacher's quality while reducing the latency to enable natural conversations.

Augmenting large language models (LLMs) to understand audio -- including non-speech sounds and non-verbal speech -- is critically important for diverse real-world applications of LLMs. In this paper, we propose Audio Flamingo, a novel audio language model with 1) strong audio understanding abilities, 2) the ability to quickly adapt to unseen tasks via in-context learning and retrieval, and 3) strong multi-turn dialogue abilities. We introduce a series of training techniques, architecture design, and data strategies to enhance our model with these abilities. Extensive evaluations across various audio understanding tasks confirm the efficacy of our method, setting new state-of-the-art benchmarks.

Speech-driven 3D facial animation is challenging due to the complex geometry of human faces and the limited availability of 3D audio-visual data. Prior works typically focus on learning phoneme-level features of short audio windows with limited context, occasionally resulting in inaccurate lip movements. To tackle this limitation, we propose a Transformer-based autoregressive model, FaceFormer, which encodes the long-term audio context and autoregressively predicts a sequence of animated 3D face meshes. To cope with the data scarcity issue, we integrate the self-supervised pre-trained speech representations. Also, we devise two biased attention mechanisms well suited to this specific task, including the biased cross-modal multi-head (MH) attention and the biased causal MH self-attention with a periodic positional encoding strategy. The former effectively aligns the audio-motion modalities, whereas the latter offers abilities to generalize to longer audio sequences. Extensive experiments and a perceptual user study show that our approach outperforms the existing state-of-the-arts. The code will be made available.

Video and audio content creation serves as the core technique for the movie industry and professional users. Recently, existing diffusion-based methods tackle video and audio generation separately, which hinders the technique transfer from academia to industry. In this work, we aim at filling the gap, with a carefully designed optimization-based framework for cross-visual-audio and joint-visual-audio generation. We observe the powerful generation ability of off-the-shelf video or audio generation models. Thus, instead of training the giant models from scratch, we propose to bridge the existing strong models with a shared latent representation space. Specifically, we propose a multimodality latent aligner with the pre-trained ImageBind model. Our latent aligner shares a similar core as the classifier guidance that guides the diffusion denoising process during inference time. Through carefully designed optimization strategy and loss functions, we show the superior performance of our method on joint video-audio generation, visual-steered audio generation, and audio-steered visual generation tasks. The project website can be found at https://yzxing87.github.io/Seeing-and-Hearing/

We introduce Generative Spoken Language Modeling, the task of learning the acoustic and linguistic characteristics of a language from raw audio (no text, no labels), and a set of metrics to automatically evaluate the learned representations at acoustic and linguistic levels for both encoding and generation. We set up baseline systems consisting of a discrete speech encoder (returning pseudo-text units), a generative language model (trained on pseudo-text), and a speech decoder (generating a waveform from pseudo-text) all trained without supervision and validate the proposed metrics with human evaluation. Across 3 speech encoders (CPC, wav2vec 2.0, HuBERT), we find that the number of discrete units (50, 100, or 200) matters in a task-dependent and encoder-dependent way, and that some combinations approach text-based systems.

Despite recent advances in syncing lip movements with any audio waves, current methods still struggle to balance generation quality and the model's generalization ability. Previous studies either require long-term data for training or produce a similar movement pattern on all subjects with low quality. In this paper, we propose StyleSync, an effective framework that enables high-fidelity lip synchronization. We identify that a style-based generator would sufficiently enable such a charming property on both one-shot and few-shot scenarios. Specifically, we design a mask-guided spatial information encoding module that preserves the details of the given face. The mouth shapes are accurately modified by audio through modulated convolutions. Moreover, our design also enables personalized lip-sync by introducing style space and generator refinement on only limited frames. Thus the identity and talking style of a target person could be accurately preserved. Extensive experiments demonstrate the effectiveness of our method in producing high-fidelity results on a variety of scenes. Resources can be found at https://hangz-nju-cuhk.github.io/projects/StyleSync.

Large-scale self-supervised pre-trained speech encoders outperform conventional approaches in speech recognition and translation tasks. Due to the high cost of developing these large models, building new encoders for new tasks and deploying them to on-device applications are infeasible. Prior studies propose model compression methods to address this issue, but those works focus on smaller models and less realistic tasks. Thus, we propose Contrastive Layer-to-layer Distillation (CoLLD), a novel knowledge distillation method to compress pre-trained speech encoders by leveraging masked prediction and contrastive learning to train student models to copy the behavior of a large teacher model. CoLLD outperforms prior methods and closes the gap between small and large models on multilingual speech-to-text translation and recognition benchmarks.

Recent TTS models with decoder-only Transformer architecture, such as SPEAR-TTS and VALL-E, achieve impressive naturalness and demonstrate the ability for zero-shot adaptation given a speech prompt. However, such decoder-only TTS models lack monotonic alignment constraints, sometimes leading to hallucination issues such as mispronunciation, word skipping and repeating. To address this limitation, we propose VALL-T, a generative Transducer model that introduces shifting relative position embeddings for input phoneme sequence, explicitly indicating the monotonic generation process while maintaining the architecture of decoder-only Transformer. Consequently, VALL-T retains the capability of prompt-based zero-shot adaptation and demonstrates better robustness against hallucinations with a relative reduction of 28.3% in the word error rate. Furthermore, the controllability of alignment in VALL-T during decoding facilitates the use of untranscribed speech prompts, even in unknown languages. It also enables the synthesis of lengthy speech by utilizing an aligned context window.

We introduce LyricWhiz, a robust, multilingual, and zero-shot automatic lyrics transcription method achieving state-of-the-art performance on various lyrics transcription datasets, even in challenging genres such as rock and metal. Our novel, training-free approach utilizes Whisper, a weakly supervised robust speech recognition model, and GPT-4, today's most performant chat-based large language model. In the proposed method, Whisper functions as the "ear" by transcribing the audio, while GPT-4 serves as the "brain," acting as an annotator with a strong performance for contextualized output selection and correction. Our experiments show that LyricWhiz significantly reduces Word Error Rate compared to existing methods in English and can effectively transcribe lyrics across multiple languages. Furthermore, we use LyricWhiz to create the first publicly available, large-scale, multilingual lyrics transcription dataset with a CC-BY-NC-SA copyright license, based on MTG-Jamendo, and offer a human-annotated subset for noise level estimation and evaluation. We anticipate that our proposed method and dataset will advance the development of multilingual lyrics transcription, a challenging and emerging task.

Language model based text-to-speech (TTS) models, like VALL-E, have gained attention for their outstanding in-context learning capability in zero-shot scenarios. Neural speech codec is a critical component of these models, which can convert speech into discrete token representations. However, excessive token sequences from the codec may negatively affect prediction accuracy and restrict the progression of Language model based TTS models. To address this issue, this paper proposes a novel neural speech codec with time-invariant codes named TiCodec. By encoding and quantizing time-invariant information into a separate code, TiCodec can reduce the amount of frame-level information that needs encoding, effectively decreasing the number of tokens as codes of speech. Furthermore, this paper introduces a time-invariant encoding consistency loss to enhance the consistency of time-invariant code within an utterance and force it to capture more global information, which can benefit the zero-shot TTS task. Experimental results demonstrate that TiCodec can not only enhance the quality of reconstruction speech with fewer tokens but also increase the similarity and naturalness, as well as reduce the word error rate of the synthesized speech by the TTS model.

Visual and auditory perception are two crucial ways humans experience the world. Text-to-video generation has made remarkable progress over the past year, but the absence of harmonious audio in generated video limits its broader applications. In this paper, we propose Semantic and Temporal Aligned Video-to-Audio (STA-V2A), an approach that enhances audio generation from videos by extracting both local temporal and global semantic video features and combining these refined video features with text as cross-modal guidance. To address the issue of information redundancy in videos, we propose an onset prediction pretext task for local temporal feature extraction and an attentive pooling module for global semantic feature extraction. To supplement the insufficient semantic information in videos, we propose a Latent Diffusion Model with Text-to-Audio priors initialization and cross-modal guidance. We also introduce Audio-Audio Align, a new metric to assess audio-temporal alignment. Subjective and objective metrics demonstrate that our method surpasses existing Video-to-Audio models in generating audio with better quality, semantic consistency, and temporal alignment. The ablation experiment validated the effectiveness of each module. Audio samples are available at https://y-ren16.github.io/STAV2A.

Voice assistants, such as Siri and Google Assistant, typically model audio and text separately, resulting in lost speech information and increased complexity. Recent efforts to address this with end-to-end Speech Large Language Models (LLMs) trained with supervised finetuning (SFT) have led to models ``forgetting" capabilities from text-only LLMs. Our work proposes an alternative paradigm for training Speech LLMs without instruction data, using the response of a text-only LLM to transcripts as self-supervision. Importantly, this process can be performed without annotated responses. We show that our Distilled Voice Assistant (DiVA) generalizes to Spoken Question Answering, Classification, and Translation. Furthermore, we show that DiVA better meets user preferences, achieving a 72\% win rate compared with state-of-the-art models like Qwen 2 Audio, despite using >100x less training compute.

The multi-codebook speech codec enables the application of large language models (LLM) in TTS but bottlenecks efficiency and robustness due to multi-sequence prediction. To avoid this obstacle, we propose Single-Codec, a single-codebook single-sequence codec, which employs a disentangled VQ-VAE to decouple speech into a time-invariant embedding and a phonetically-rich discrete sequence. Furthermore, the encoder is enhanced with 1) contextual modeling with a BLSTM module to exploit the temporal information, 2) a hybrid sampling module to alleviate distortion from upsampling and downsampling, and 3) a resampling module to encourage discrete units to carry more phonetic information. Compared with multi-codebook codecs, e.g., EnCodec and TiCodec, Single-Codec demonstrates higher reconstruction quality with a lower bandwidth of only 304bps. The effectiveness of Single-Code is further validated by LLM-TTS experiments, showing improved naturalness and intelligibility.

Large Language Models (LLMs) have made significant progress in various downstream tasks, inspiring the development of Speech Understanding Language Models (SULMs) to enable comprehensive speech-based interactions. However, most advanced SULMs are developed by the industry, leveraging large-scale datasets and computational resources that are not readily available to the academic community. Moreover, the lack of transparency in training details creates additional barriers to further innovation. In this study, we present OSUM, an Open Speech Understanding Model designed to explore the potential of training SLUMs under constrained academic resources. The OSUM model combines a Whisper encoder with a Qwen2 LLM and supports a wide range of speech tasks, including speech recognition (ASR), speech recognition with timestamps (SRWT), vocal event detection (VED), speech emotion recognition (SER), speaking style recognition (SSR), speaker gender classification (SGC), speaker age prediction (SAP), and speech-to-text chat (STTC). By employing an ASR+X training strategy, OSUM achieves efficient and stable multi-task training by simultaneously optimizing ASR alongside target tasks. Beyond delivering strong performance, OSUM emphasizes transparency by providing openly available data preparation and training methodologies, offering valuable insights and practical guidance for the academic community. By doing so, we aim to accelerate research and innovation in advanced SULM technologies.

Audio-Language Models (ALMs), which are trained on audio-text data, focus on the processing, understanding, and reasoning of sounds. Unlike traditional supervised learning approaches learning from predefined labels, ALMs utilize natural language as a supervision signal, which is more suitable for describing complex real-world audio recordings. ALMs demonstrate strong zero-shot capabilities and can be flexibly adapted to diverse downstream tasks. These strengths not only enhance the accuracy and generalization of audio processing tasks but also promote the development of models that more closely resemble human auditory perception and comprehension. Recent advances in ALMs have positioned them at the forefront of computer audition research, inspiring a surge of efforts to advance ALM technologies. Despite rapid progress in the field of ALMs, there is still a notable lack of systematic surveys that comprehensively organize and analyze developments. In this paper, we present a comprehensive review of ALMs with a focus on general audio tasks, aiming to fill this gap by providing a structured and holistic overview of ALMs. Specifically, we cover: (1) the background of computer audition and audio-language models; (2) the foundational aspects of ALMs, including prevalent network architectures, training objectives, and evaluation methods; (3) foundational pre-training and audio-language pre-training approaches; (4) task-specific fine-tuning, multi-task tuning and agent systems for downstream applications; (5) datasets and benchmarks; and (6) current challenges and future directions. Our review provides a clear technical roadmap for researchers to understand the development and future trends of existing technologies, offering valuable references for implementation in real-world scenarios.

In recent years, self-supervised pre-training methods have gained significant traction in learning high-level information from raw speech. Among these methods, HuBERT has demonstrated SOTA performance in automatic speech recognition (ASR). However, HuBERT's performance lags behind data2vec due to disparities in pre-training strategies. In this paper, we propose (i) a Swap method to address pre-training and inference mismatch observed in HuBERT and (ii) incorporates Multicluster masked prediction loss for more effective utilization of the models capacity. The resulting method is, MS-HuBERT, an end-to-end self-supervised pre-training method for learning robust speech representations. It beats vanilla HuBERT on the ASR Librispeech benchmark on average by a 5% margin when evaluated on different finetuning splits. Additionally, we demonstrate that the learned embeddings obtained during pre-training encode essential information for improving performance of content based tasks such as ASR.

By leveraging the power of Large Language Models(LLMs) and speech foundation models, state of the art speech-text bimodal works can achieve challenging tasks like spoken translation(ST) and question answering(SQA) altogether with much simpler architectures. In this paper, we utilize the capability of Whisper encoder and pre-trained Yi-6B. Empirical results reveal that modal alignment can be achieved with one layer module and hundred hours of speech-text multitask corpus. We further swap the Yi-6B with human preferences aligned version of Yi-6B-Chat during inference, and discover that the alignment capability is applicable as well. In addition, the alignment subspace revealed by singular value decomposition(SVD) also implies linear alignment subspace is sparse, which leaves the possibility to concatenate other features like voice-print or video to expand modality.

Recent advances in audio generation have focused on text-to-audio (T2A) and video-to-audio (V2A) tasks. However, T2A or V2A methods cannot generate holistic sounds (onscreen and off-screen). This is because T2A cannot generate sounds aligning with onscreen objects, while V2A cannot generate semantically complete (offscreen sounds missing). In this work, we address the task of holistic audio generation: given a video and a text prompt, we aim to generate both onscreen and offscreen sounds that are temporally synchronized with the video and semantically aligned with text and video. Previous approaches for joint text and video-to-audio generation often suffer from modality bias, favoring one modality over the other. To overcome this limitation, we introduce VinTAGe, a flow-based transformer model that jointly considers text and video to guide audio generation. Our framework comprises two key components: a Visual-Text Encoder and a Joint VT-SiT model. To reduce modality bias and improve generation quality, we employ pretrained uni-modal text-to-audio and video-to-audio generation models for additional guidance. Due to the lack of appropriate benchmarks, we also introduce VinTAGe-Bench, a dataset of 636 video-text-audio pairs containing both onscreen and offscreen sounds. Our comprehensive experiments on VinTAGe-Bench demonstrate that joint text and visual interaction is necessary for holistic audio generation. Furthermore, VinTAGe achieves state-of-the-art results on the VGGSound benchmark. Our source code and pre-trained models will be released. Demo is available at: https://www.youtube.com/watch?v=QmqWhUjPkJI.

Efficiently compressing high-dimensional audio signals into a compact and informative latent space is crucial for various tasks, including generative modeling and music information retrieval (MIR). Existing audio autoencoders, however, often struggle to achieve high compression ratios while preserving audio fidelity and facilitating efficient downstream applications. We introduce Music2Latent2, a novel audio autoencoder that addresses these limitations by leveraging consistency models and a novel approach to representation learning based on unordered latent embeddings, which we call summary embeddings. Unlike conventional methods that encode local audio features into ordered sequences, Music2Latent2 compresses audio signals into sets of summary embeddings, where each embedding can capture distinct global features of the input sample. This enables to achieve higher reconstruction quality at the same compression ratio. To handle arbitrary audio lengths, Music2Latent2 employs an autoregressive consistency model trained on two consecutive audio chunks with causal masking, ensuring coherent reconstruction across segment boundaries. Additionally, we propose a novel two-step decoding procedure that leverages the denoising capabilities of consistency models to further refine the generated audio at no additional cost. Our experiments demonstrate that Music2Latent2 outperforms existing continuous audio autoencoders regarding audio quality and performance on downstream tasks. Music2Latent2 paves the way for new possibilities in audio compression.

Sample-and-rank is a key decoding strategy for modern generation-based dialogue systems. It helps achieve diverse and high-quality responses by selecting an answer from a small pool of generated candidates. The current state-of-the-art ranking methods mainly use an encoding paradigm called Cross-Encoder, which separately encodes each context-candidate pair and ranks the candidates according to their fitness scores. However, Cross-Encoder repeatedly encodes the same lengthy context for each candidate, resulting in high computational costs. Poly-Encoder addresses the above problems by reducing the interaction between context and candidates, but with a price of performance drop. In this work, we develop a new paradigm called Uni-Encoder, that keeps the full attention over each pair as in Cross-Encoder while only encoding the context once, as in Poly-Encoder. Uni-Encoder encodes all the candidates with the context in one forward pass. We use the same positional embedding for all candidates to ensure they are treated equally and design a new attention mechanism to avoid confusion. Our Uni-Encoder can simulate other ranking paradigms using different attention and response concatenation methods. Extensive experiments show that our proposed paradigm achieves new state-of-the-art results on four benchmark datasets with high computational efficiency. For instance, it improves R10@1 by 2.9% with an approximately 4X faster inference speed on the Ubuntu V2 dataset.

One-shot voice conversion aims to change the timbre of any source speech to match that of the unseen target speaker with only one speech sample. Existing methods face difficulties in satisfactory speech representation disentanglement and suffer from sizable networks as some of them leverage numerous complex modules for disentanglement. In this paper, we propose a model named MAIN-VC to effectively disentangle via a concise neural network. The proposed model utilizes Siamese encoders to learn clean representations, further enhanced by the designed mutual information estimator. The Siamese structure and the newly designed convolution module contribute to the lightweight of our model while ensuring performance in diverse voice conversion tasks. The experimental results show that the proposed model achieves comparable subjective scores and exhibits improvements in objective metrics compared to existing methods in a one-shot voice conversion scenario.

We propose Cotatron, a transcription-guided speech encoder for speaker-independent linguistic representation. Cotatron is based on the multispeaker TTS architecture and can be trained with conventional TTS datasets. We train a voice conversion system to reconstruct speech with Cotatron features, which is similar to the previous methods based on Phonetic Posteriorgram (PPG). By training and evaluating our system with 108 speakers from the VCTK dataset, we outperform the previous method in terms of both naturalness and speaker similarity. Our system can also convert speech from speakers that are unseen during training, and utilize ASR to automate the transcription with minimal reduction of the performance. Audio samples are available at https://mindslab-ai.github.io/cotatron, and the code with a pre-trained model will be made available soon.

Data scarcity and the modality gap between the speech and text modalities are two major obstacles of end-to-end Speech Translation (ST) systems, thus hindering their performance. Prior work has attempted to mitigate these challenges by leveraging external MT data and optimizing distance metrics that bring closer the speech-text representations. However, achieving competitive results typically requires some ST data. For this reason, we introduce ZeroSwot, a method for zero-shot ST that bridges the modality gap without any paired ST data. Leveraging a novel CTC compression and Optimal Transport, we train a speech encoder using only ASR data, to align with the representation space of a massively multilingual MT model. The speech encoder seamlessly integrates with the MT model at inference, enabling direct translation from speech to text, across all languages supported by the MT model. Our experiments show that we can effectively close the modality gap without ST data, while our results on MuST-C and CoVoST demonstrate our method's superiority over not only previous zero-shot models, but also supervised ones, achieving state-of-the-art results.

Recent advancements in audio generation have enabled the creation of high-fidelity audio clips from free-form textual descriptions. However, temporal relationships, a critical feature for audio content, are currently underrepresented in mainstream models, resulting in an imprecise temporal controllability. Specifically, users cannot accurately control the timestamps of sound events using free-form text. We acknowledge that a significant factor is the absence of high-quality, temporally-aligned audio-text datasets, which are essential for training models with temporal control. The more temporally-aligned the annotations, the better the models can understand the precise relationship between audio outputs and temporal textual prompts. Therefore, we present a strongly aligned audio-text dataset, AudioTime. It provides text annotations rich in temporal information such as timestamps, duration, frequency, and ordering, covering almost all aspects of temporal control. Additionally, we offer a comprehensive test set and evaluation metric to assess the temporal control performance of various models. Examples are available on the https://zeyuxie29.github.io/AudioTime/

Recently, various studies have been directed towards exploring dense passage retrieval techniques employing pre-trained language models, among which the masked auto-encoder (MAE) pre-training architecture has emerged as the most promising. The conventional MAE framework relies on leveraging the passage reconstruction of decoder to bolster the text representation ability of encoder, thereby enhancing the performance of resulting dense retrieval systems. Within the context of building the representation ability of the encoder through passage reconstruction of decoder, it is reasonable to postulate that a ``more demanding'' decoder will necessitate a corresponding increase in the encoder's ability. To this end, we propose a novel token importance aware masking strategy based on pointwise mutual information to intensify the challenge of the decoder. Importantly, our approach can be implemented in an unsupervised manner, without adding additional expenses to the pre-training phase. Our experiments verify that the proposed method is both effective and robust on large-scale supervised passage retrieval datasets and out-of-domain zero-shot retrieval benchmarks.

We describe a method to jointly pre-train speech and text in an encoder-decoder modeling framework for speech translation and recognition. The proposed method incorporates four self-supervised and supervised subtasks for cross modality learning. A self-supervised speech subtask leverages unlabelled speech data, and a (self-)supervised text to text subtask makes use of abundant text training data. Two auxiliary supervised speech tasks are included to unify speech and text modeling space. Our contribution lies in integrating linguistic information from the text corpus into the speech pre-training. Detailed analysis reveals learning interference among subtasks. Two pre-training configurations for speech translation and recognition, respectively, are presented to alleviate subtask interference. Our experiments show the proposed method can effectively fuse speech and text information into one model. It achieves between 1.7 and 2.3 BLEU improvement above the state of the art on the MuST-C speech translation dataset and comparable WERs to wav2vec 2.0 on the Librispeech speech recognition task.

Although audio generation shares commonalities across different types of audio, such as speech, music, and sound effects, designing models for each type requires careful consideration of specific objectives and biases that can significantly differ from those of other types. To bring us closer to a unified perspective of audio generation, this paper proposes a framework that utilizes the same learning method for speech, music, and sound effect generation. Our framework introduces a general representation of audio, called language of audio (LOA). Any audio can be translated into LOA based on AudioMAE, a self-supervised pre-trained representation learning model. In the generation process, we translate any modalities into LOA by using a GPT-2 model, and we perform self-supervised audio generation learning with a latent diffusion model conditioned on LOA. The proposed framework naturally brings advantages such as in-context learning abilities and reusable self-supervised pretrained AudioMAE and latent diffusion models. Experiments on the major benchmarks of text-to-audio, text-to-music, and text-to-speech demonstrate new state-of-the-art or competitive performance to previous approaches. Our demo and code are available at https://audioldm.github.io/audioldm2.

Voice conversion refers to transferring speaker identity with well-preserved content. Better disentanglement of speech representations leads to better voice conversion. Recent studies have found that phonetic information from input audio has the potential ability to well represent content. Besides, the speaker-style modeling with pre-trained models making the process more complex. To tackle these issues, we introduce a new method named "CTVC" which utilizes disentangled speech representations with contrastive learning and time-invariant retrieval. Specifically, a similarity-based compression module is used to facilitate a more intimate connection between the frame-level hidden features and linguistic information at phoneme-level. Additionally, a time-invariant retrieval is proposed for timbre extraction based on multiple segmentations and mutual information. Experimental results demonstrate that "CTVC" outperforms previous studies and improves the sound quality and similarity of converted results.

Distributional shift is a central challenge in the deployment of machine learning models as they can be ill-equipped for real-world data. This is particularly evident in text-to-audio generation where the encoded representations are easily undermined by unseen prompts, which leads to the degradation of generated audio -- the limited set of the text-audio pairs remains inadequate for conditional audio generation in the wild as user prompts are under-specified. In particular, we observe a consistent audio quality degradation in generated audio samples with user prompts, as opposed to training set prompts. To this end, we present a retrieval-based in-context prompt editing framework that leverages the training captions as demonstrative exemplars to revisit the user prompts. We show that the framework enhanced the audio quality across the set of collected user prompts, which were edited with reference to the training captions as exemplars.