EuroSpeech Dataset

Dataset Description

EuroSpeech is a large-scale multilingual speech corpus containing high-quality aligned parliamentary speech across 18 European languages. The dataset was constructed by processing parliamentary proceedings using a robust alignment pipeline that handles diverse audio formats and non-verbatim transcripts.

Dataset Summary

• Languages: 18 European languages (see detailed breakdown below)
• Total aligned hours: ~58,200 hours of aligned speech-text data
• Quality-filtered subsets:
  • CER < 30%: approximately 48,350 hours (78.2% of aligned data)
  • CER < 20%: approximately 40,950 hours (66.3% of aligned data) - primary subset
  • CER < 10%: approximately 28,457 hours (46.0% of aligned data)
• Domain: Parliamentary proceedings (formal speaking style)
• Audio segment length: Typically 3-20 seconds
• Format: Audio segments with paired transcriptions

Languages

EuroSpeech provides substantial data for previously under-resourced languages:

• 15 languages exceed 1,000 hours of data (CER < 20%)
• 17 languages exceed 500 hours of data (CER < 20%)

Language	Code	Total Aligned (h)	CER < 30% (h)	CER < 20% (h)	CER < 10% (h)
Croatia	hr	7485.8	5899.7	5560.3	4592.0
Denmark	da	7014.6	6435.0	5415.4	3443.7
Norway	no	5327.4	4578.8	3746.2	2252.2
Portugal	pt	5096.8	4036.7	3196.8	2105.9
Italy	it	4815.5	3539.6	2724.0	1767.3
Slovakia	sk	2820.9	2679.6	2524.8	2037.9
Greece	el	3096.9	2717.6	2341.9	1620.9
Sweden	sv	3849.8	2862.6	2231.2	1360.1
Germany	de	2472.8	2354.2	2154.7	1698.4
Bulgaria	bg	3420.7	2570.4	2146.2	1472.8
Finland	fi	2131.2	1991.4	1823.0	1442.2
Serbia	sr	2263.6	1985.1	1830.1	1374.1
Ukraine	uk	1287.9	1238.3	1182.6	1029.8
Slovenia	sl	1338.9	1241.7	1141.6	900.5
Latvia	lv	2049.6	1627.9	1158.9	499.9
Estonia	et	1378.9	1235.0	856.3	351.9
Bosnia & Herzegovina	bcs	855.5	777.5	675.7	445.3
Malta	mt	1486.2	579.1	244.4	61.9
Total		58193.0	48350.2	40954.7	28456.8

Dataset Structure

Data Instances

Each instance in the dataset consists of:

• Audio segment (3-20 seconds)
• Corresponding transcript text
• Metadata including language, source session, alignment quality metrics

Data Splits

The dataset provides predefined train, development, and test splits for each language. To ensure data integrity and prevent leakage between sets, these splits are constructed by assigning entire parliamentary sessions (i.e., all segments derived from a single original long audio recording) exclusively to one of the train, development, or test sets. The exact proportions follow common practices (e.g., 80/10/10).

Dataset Creation

Source Data

The data was collected from parliamentary proceedings across 18 European nations. Parliamentary sessions offer high-quality speech in a formal register, typically featuring clear speech with good audio quality and professional transcripts.

Data Collection and Processing

The dataset was constructed using a multi-stage pipeline:

1. Data Sourcing and Metadata Collection: Manual and scripted gathering of media/transcript links from parliamentary websites.

2. Download Pipeline: Automated retrieval of audio, video, and transcript files using specialized handlers for diverse source formats.

3. Alignment Pipeline:

   • Segmentation of long recordings into 3-20 second utterances using voice activity detection (VAD)
   • Transcription of segments using an ASR model to produce pseudo-labels
   • Alignment of segments to transcripts using a novel two-stage dynamic algorithm
   • Selection of best-aligned transcript formats and quality filtering

4. Filtering: CER-based filtering to create quality tiers (CER < 30%, < 20%, < 10%)

Alignment Algorithm

The core of the alignment process is a novel two-stage dynamic algorithm specifically engineered for extreme robustness when matching ASR pseudo-labels to noisy, non-verbatim parliamentary transcripts:

1. Coarse stage: Uses a sliding window to rapidly scan the transcript, efficiently bypassing large irrelevant sections to identify a set of top-k candidate text spans via Character Error Rate (CER).

2. Fine-tuning stage: Performs a local search around promising candidates, optimizing start position and window size for the best CER.

A fallback mechanism restarts the search if no initial match meets a predefined quality threshold.

Dataset Use

Intended Uses

The EuroSpeech dataset is intended for:

• Training and evaluating automatic speech recognition (ASR) systems
• Training and evaluating text-to-speech (TTS) systems
• Multilingual speech research
• Low-resource language speech technology development
• Cross-lingual transfer learning in speech models

Citation Information

If you use this dataset, please cite:

[Citation details to be added upon publication]

Considerations

Data Quality

The dataset provides multiple quality tiers based on Character Error Rate (CER):

• CER < 30%: More data, but potentially lower quality alignments
• CER < 20%: Balanced quality-quantity trade-off (recommended for most applications)
• CER < 10%: Highest quality alignments, but reduced quantity

Licensing Information

[License details to be added]

Limitations

• The dataset primarily represents formal parliamentary speech and may not generalize well to casual, spontaneous, or noisy speech environments.
• The dataset reflects the demographics and speaking styles of European parliamentarians, which may not be representative of the general population.
• Some languages have significantly more data than others, which could lead to performance disparities in multilingual models.

Additional Information

Dataset Curators

[To be completed with author information after publication]

Maintenance Status

[Information about maintenance and update plans]

Links

• [Download links to be added upon publication]
• [GitHub repository link to be added]