Datasets:
Tasks:
Visual Question Answering
Sub-tasks:
visual-question-answering
Languages:
Chinese
Size:
< 1K
License:
Upload folder using huggingface_hub
Browse filesThis view is limited to 50 files because it contains too many changes.
See raw diff
- README.md +13 -1
- README_CN.md +27 -15
- imgs/07_collage.jpg +2 -2
- media/01/1_1_1.mp4 +3 -0
- media/01/1_1_10.mp4 +3 -0
- media/01/1_1_11.mp4 +3 -0
- media/01/1_1_12.mp4 +3 -0
- media/01/1_1_13.mp4 +3 -0
- media/01/1_1_14.mp4 +3 -0
- media/01/1_1_15.mp4 +3 -0
- media/01/1_1_16.mp4 +3 -0
- media/01/1_1_17.mp4 +3 -0
- media/01/1_1_18.mp4 +3 -0
- media/01/1_1_19.mp4 +3 -0
- media/01/1_1_2.mp4 +3 -0
- media/01/1_1_20.mp4 +3 -0
- media/01/1_1_21.mp4 +3 -0
- media/01/1_1_22.mp4 +3 -0
- media/01/1_1_23.mp4 +3 -0
- media/01/1_1_24.mp4 +3 -0
- media/01/1_1_25.jpg +3 -0
- media/01/1_1_26.jpg +3 -0
- media/01/1_1_27.jpg +3 -0
- media/01/1_1_28.jpg +3 -0
- media/01/1_1_29.jpg +3 -0
- media/01/1_1_3.mp4 +3 -0
- media/01/1_1_30.jpg +3 -0
- media/01/1_1_31.jpg +3 -0
- media/01/1_1_32.jpeg +3 -0
- media/01/1_1_33.jpg +3 -0
- media/01/1_1_34.jpeg +3 -0
- media/01/1_1_35.jpeg +3 -0
- media/01/1_1_36.png +3 -0
- media/01/1_1_37.png +3 -0
- media/01/1_1_38.png +3 -0
- media/01/1_1_39.png +3 -0
- media/01/1_1_4.mp4 +3 -0
- media/01/1_1_40.png +3 -0
- media/01/1_1_41.png +3 -0
- media/01/1_1_42.png +3 -0
- media/01/1_1_43.png +3 -0
- media/01/1_1_44.png +3 -0
- media/01/1_1_45.png +3 -0
- media/01/1_1_46.png +3 -0
- media/01/1_1_47.png +3 -0
- media/01/1_1_48.png +3 -0
- media/01/1_1_49.png +3 -0
- media/01/1_1_5.mp4 +3 -0
- media/01/1_1_50.jpg +3 -0
- media/01/1_1_6.mp4 +3 -0
README.md
CHANGED
@@ -4,6 +4,8 @@
|
|
4 |
|
5 |
**The Titan CV Benchmark is primarily designed to evaluate the performance of AGENTS in the field of computer vision (CV).** We have collected more than 200 test examples to comprehensively test the performance of the agent, particularly their capability to solve problems step by step.
|
6 |
|
|
|
|
|
7 |
|
8 |
|
9 |
## Basic principles
|
@@ -20,6 +22,16 @@ In order to distinguish the traditional benchmark of VLM, all collected samples
|
|
20 |
|
21 |
|
22 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
23 |
## Quantity Classification
|
24 |
|
25 |
This test set collects most of the CV problems on the market and divides them into **4 categories and 22 subcategories**:
|
@@ -104,7 +116,7 @@ This test set collects most of the CV problems on the market and divides them in
|
|
104 |
|
105 |
Each data item contains 6 fields:
|
106 |
|
107 |
-
```
|
108 |
"id": Data item number, named according to "categories_subcategories_items".
|
109 |
"media_path": The path to the media file associated with the data item.
|
110 |
"media_type": Media type that the data item depends on.
|
|
|
4 |
|
5 |
**The Titan CV Benchmark is primarily designed to evaluate the performance of AGENTS in the field of computer vision (CV).** We have collected more than 200 test examples to comprehensively test the performance of the agent, particularly their capability to solve problems step by step.
|
6 |
|
7 |
+
**Demo Video: https://youtu.be/dcUb4lUnGj4**
|
8 |
+
|
9 |
|
10 |
|
11 |
## Basic principles
|
|
|
22 |
|
23 |
|
24 |
|
25 |
+
## Titan CV Agent Series
|
26 |
+
|
27 |
+
[Titan CV Agent Sandbox](https://github.com/DataCanvasAILab/Titan-CV-Agent-Sandbox)
|
28 |
+
|
29 |
+
[Titan CV Agent Executor](https://github.com/DataCanvasAILab/Titan-CV-Agent-Executor)
|
30 |
+
|
31 |
+
[Titan CV Agent Benchmark](https://github.com/DataCanvasAILab/Titan-CV-Agent-Benchmark)
|
32 |
+
|
33 |
+
|
34 |
+
|
35 |
## Quantity Classification
|
36 |
|
37 |
This test set collects most of the CV problems on the market and divides them into **4 categories and 22 subcategories**:
|
|
|
116 |
|
117 |
Each data item contains 6 fields:
|
118 |
|
119 |
+
```
|
120 |
"id": Data item number, named according to "categories_subcategories_items".
|
121 |
"media_path": The path to the media file associated with the data item.
|
122 |
"media_type": Media type that the data item depends on.
|
README_CN.md
CHANGED
@@ -1,6 +1,8 @@
|
|
1 |
# Titan 计算机视觉智能体评测基准
|
2 |
|
3 |
-
Titan
|
|
|
|
|
4 |
|
5 |
|
6 |
|
@@ -8,13 +10,23 @@ Titan 计算机视觉智能体评测基准主要用于测试CV领域的Agent性
|
|
8 |
|
9 |
为了区别传统测试视觉-语言模型的评测基准,所有收集的样例将遵循以下原则:
|
10 |
|
11 |
-
①
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
12 |
|
13 |
-
② 每个问题将尽力维持其复杂性,不是任何一步(例如只使用VLM)就能够解决的。而是需要多步骤的串行思考。
|
14 |
|
15 |
-
|
16 |
|
17 |
-
|
|
|
|
|
|
|
|
|
18 |
|
19 |
|
20 |
|
@@ -98,7 +110,7 @@ Titan 计算机视觉智能体评测基准主要用于测试CV领域的Agent性
|
|
98 |
|
99 |
每个数据项包含6个字段:
|
100 |
|
101 |
-
```
|
102 |
"id": 数据项编号,按照“大类_小类_标号”命名。
|
103 |
"media_path": 数据项依赖的媒体位置。
|
104 |
"media_type": 数据项依赖的媒体类型。
|
@@ -176,24 +188,24 @@ Titan 计算机视觉智能体评测基准主要用于测试CV领域的Agent性
|
|
176 |
|
177 |
|
178 |
|
179 |
-
##
|
180 |
|
181 |
-
|
182 |
|
183 |
-
***[
|
184 |
|
185 |
-
>
|
186 |
>
|
187 |
-
>
|
188 |
>
|
189 |
-
>
|
190 |
>
|
191 |
-
>
|
192 |
|
193 |
|
194 |
|
195 |
-
##
|
196 |
|
197 |
-
Titan
|
198 |
|
199 |
**Apache License Version 2.0**
|
|
|
1 |
# Titan 计算机视觉智能体评测基准
|
2 |
|
3 |
+
Titan 计算机视觉智能体评测基准主要用于测试计算机视觉领域的智能体性能。我们广泛收集了200多条测试样例全面测试智能体的性能,尤其是其分步骤解决问题的能力。
|
4 |
+
|
5 |
+
**演示视频:https://youtu.be/dcUb4lUnGj4**
|
6 |
|
7 |
|
8 |
|
|
|
10 |
|
11 |
为了区别传统测试视觉-语言模型的评测基准,所有收集的样例将遵循以下原则:
|
12 |
|
13 |
+
① **模态多样**,包括各种类型的图片和视频。领域丰富,包含工业、医疗、农业、环境、社交、文化、体育以及科学研究等多种领域。
|
14 |
+
|
15 |
+
② 每个问题将尽力维持其复杂性,不是任何一步(例如只使用VLM)就能够解决的。而是需要**多步骤的串行思考**。
|
16 |
+
|
17 |
+
③ 每个问题将**专注考察智能体的智能**,它必须对人类简单、对机器复杂,它看起来不复杂,除了医学领域,可以被任何一个毫无经验的成年人轻易理解,但需要充分调动其思考、认知能力。对于机器来说,传统的背诵、计算、速度优势是无益于解出这些题目的,需要其与人类相似的空间认知或高级理解能力。
|
18 |
+
|
19 |
+
④ 每个问题是复杂的,但**答案是简单的、可验证的**,避免开放式问答对结果检验的压力,如果类似位置区别的问题,为避免歧义,会给予一定的选项,但同时为了避免智能体猜测到正确答案,备选答案通常会设置到6个以上,降低其猜测的风险。
|
20 |
+
|
21 |
|
|
|
22 |
|
23 |
+
## Titan 计算机视觉智能体系列
|
24 |
|
25 |
+
[Titan 计算机视觉智能体沙箱](https://github.com/DataCanvasAILab/Titan-CV-Agent-Sandbox)
|
26 |
+
|
27 |
+
[Titan 计算机视觉智能体执行器](https://github.com/DataCanvasAILab/Titan-CV-Agent-Executor)
|
28 |
+
|
29 |
+
[Titan 计算机视觉智能体评测基准](https://github.com/DataCanvasAILab/Titan-CV-Agent-Benchmark)
|
30 |
|
31 |
|
32 |
|
|
|
110 |
|
111 |
每个数据项包含6个字段:
|
112 |
|
113 |
+
```
|
114 |
"id": 数据项编号,按照“大类_小类_标号”命名。
|
115 |
"media_path": 数据项依赖的媒体位置。
|
116 |
"media_type": 数据项依赖的媒体类型。
|
|
|
188 |
|
189 |
|
190 |
|
191 |
+
## 致谢
|
192 |
|
193 |
+
此项目由北京九章云极科技有限公司提供支持:
|
194 |
|
195 |
+
***[九章云极 - 助力全球企业智能升级](https://www.datacanvas.com/AboutUs)***
|
196 |
|
197 |
+
> 咨询热线:+86 400-805-7188
|
198 |
>
|
199 |
+
> 电子邮件:[email protected]
|
200 |
>
|
201 |
+
> 北京职场:北京市海淀区和盈中心C座6层&8层
|
202 |
>
|
203 |
+
> 上海职场:上海市静安区市北壹中心1号楼10层
|
204 |
|
205 |
|
206 |
|
207 |
+
## 许可证
|
208 |
|
209 |
+
Titan 计算机视觉智能体评测基准将使用下列许可证:
|
210 |
|
211 |
**Apache License Version 2.0**
|
imgs/07_collage.jpg
CHANGED
![]() |
Git LFS Details
|
![]() |
Git LFS Details
|
media/01/1_1_1.mp4
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c54a9255a27889be1c4e36a00b78c88f86abba67d19e5d5e1dccb11e6ca44c9a
|
3 |
+
size 9402794
|
media/01/1_1_10.mp4
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:8cf3998da2a2e393a1b79e5d6a5bb79129232986066ae8fc5393b49436d8eee7
|
3 |
+
size 8171238
|
media/01/1_1_11.mp4
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:85e87e9e8e62e3b5557adb63eb857b39ddd644474bc0936552032d10d7e6bcfe
|
3 |
+
size 15199347
|
media/01/1_1_12.mp4
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:b995042f1e0f3c834ed52cc77daa04dcdac5638ff6d6424b04bf38d8be855f57
|
3 |
+
size 7998902
|
media/01/1_1_13.mp4
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:3345a2a3076383dde5e6b337c3d0e163713f246cdc1c61f99815f478f401f97c
|
3 |
+
size 2988074
|
media/01/1_1_14.mp4
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:8ae930ca36141804db6a0df5a1d187f08ae7067a5fd05097952f4dc522128348
|
3 |
+
size 5395019
|
media/01/1_1_15.mp4
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:35150ba8e9b08970258c54ffc9c110e8d554b34f0924c365a5da506af7bf126b
|
3 |
+
size 3762217
|
media/01/1_1_16.mp4
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:88999b82cbcf34ae02f18d5c36bc652423878a83cb07b20be277a22a0df791e3
|
3 |
+
size 25608671
|
media/01/1_1_17.mp4
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:e3a57bdae3deba574e759fd93cd3bfc41f5af8f396938e6cd9bd3ee63da3883e
|
3 |
+
size 10621470
|
media/01/1_1_18.mp4
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:04dffec02313e4a29f7f66782abaac5b9b1918cbe5ca80edbc7ebc65c484b3d0
|
3 |
+
size 15945180
|
media/01/1_1_19.mp4
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:d4bffc17efbe4531b4498ecb037f28cafdb33f42711b6d4c19d1817481717433
|
3 |
+
size 4301151
|
media/01/1_1_2.mp4
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:54fb8ca0bd0f818ab9b25ec4e128c239219a117dd309b32b6fd9c66d7a36852d
|
3 |
+
size 45102617
|
media/01/1_1_20.mp4
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:a44c09354c9a802b73633b73c8cdace63a566ec17287ec6b056e4b4f4c1ba4dd
|
3 |
+
size 3921469
|
media/01/1_1_21.mp4
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:635f0ab66fce46db3e984e6390747f9ebc3e4badb57805e373486e61d4b83145
|
3 |
+
size 6582475
|
media/01/1_1_22.mp4
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:1f8a5c1645400d0e9374ee8c6f86c0a6393db7342fd97a9d6de048f3416d022e
|
3 |
+
size 503324
|
media/01/1_1_23.mp4
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:5c48c2e58018b4f26f0677a71163da600f269aa73784a11b27cfe99c5f7d42b5
|
3 |
+
size 1218915
|
media/01/1_1_24.mp4
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:87022b9b75bc837dc349fd7c7c365a11057d190778fc8245b1bb6764b5a21820
|
3 |
+
size 1902335
|
media/01/1_1_25.jpg
ADDED
![]() |
Git LFS Details
|
media/01/1_1_26.jpg
ADDED
![]() |
Git LFS Details
|
media/01/1_1_27.jpg
ADDED
![]() |
Git LFS Details
|
media/01/1_1_28.jpg
ADDED
![]() |
Git LFS Details
|
media/01/1_1_29.jpg
ADDED
![]() |
Git LFS Details
|
media/01/1_1_3.mp4
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:be2cfe8b79670e5ea1c5cf4d37c9b66c22d0185cea4b6c2cce6b6bebb6e6a1e9
|
3 |
+
size 17462550
|
media/01/1_1_30.jpg
ADDED
![]() |
Git LFS Details
|
media/01/1_1_31.jpg
ADDED
![]() |
Git LFS Details
|
media/01/1_1_32.jpeg
ADDED
![]() |
Git LFS Details
|
media/01/1_1_33.jpg
ADDED
![]() |
Git LFS Details
|
media/01/1_1_34.jpeg
ADDED
![]() |
Git LFS Details
|
media/01/1_1_35.jpeg
ADDED
![]() |
Git LFS Details
|
media/01/1_1_36.png
ADDED
![]() |
Git LFS Details
|
media/01/1_1_37.png
ADDED
![]() |
Git LFS Details
|
media/01/1_1_38.png
ADDED
![]() |
Git LFS Details
|
media/01/1_1_39.png
ADDED
![]() |
Git LFS Details
|
media/01/1_1_4.mp4
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:5183519534b1e861bf5e9958832848c86e8966e21bef7a33d23256612eed7de5
|
3 |
+
size 29042449
|
media/01/1_1_40.png
ADDED
![]() |
Git LFS Details
|
media/01/1_1_41.png
ADDED
![]() |
Git LFS Details
|
media/01/1_1_42.png
ADDED
![]() |
Git LFS Details
|
media/01/1_1_43.png
ADDED
![]() |
Git LFS Details
|
media/01/1_1_44.png
ADDED
![]() |
Git LFS Details
|
media/01/1_1_45.png
ADDED
![]() |
Git LFS Details
|
media/01/1_1_46.png
ADDED
![]() |
Git LFS Details
|
media/01/1_1_47.png
ADDED
![]() |
Git LFS Details
|
media/01/1_1_48.png
ADDED
![]() |
Git LFS Details
|
media/01/1_1_49.png
ADDED
![]() |
Git LFS Details
|
media/01/1_1_5.mp4
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:a967e5f7d351ead4a1905c67a4e831843a657f2ddfbc204b599aebe31a1cfc35
|
3 |
+
size 12344263
|
media/01/1_1_50.jpg
ADDED
![]() |
Git LFS Details
|
media/01/1_1_6.mp4
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:3e5bcc0d7efc9e70baca20faa2bc776c7ef1d645846987712eb5edfee810528d
|
3 |
+
size 7518687
|