MMVU: Measuring Expert-Level Multi-Discipline Video Understanding
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866917898637803520 |
|---|---|
| author | Zhao, Yilun Xie, Lujing Zhang, Haowei Gan, Guo Long, Yitao Hu, Zhiyuan Hu, Tongyan Chen, Weiyuan Li, Chuhan Song, Junyang Xu, Zhijian Wang, Chengye Pan, Weifeng Shangguan, Ziyao Tang, Xiangru Liang, Zhenwen Liu, Yixin Zhao, Chen Cohan, Arman |
| author_facet | Zhao, Yilun Xie, Lujing Zhang, Haowei Gan, Guo Long, Yitao Hu, Zhiyuan Hu, Tongyan Chen, Weiyuan Li, Chuhan Song, Junyang Xu, Zhijian Wang, Chengye Pan, Weifeng Shangguan, Ziyao Tang, Xiangru Liang, Zhenwen Liu, Yixin Zhao, Chen Cohan, Arman |
| contents | We introduce MMVU, a comprehensive expert-level, multi-discipline benchmark for evaluating foundation models in video understanding. MMVU includes 3,000 expert-annotated questions spanning 27 subjects across four core disciplines: Science, Healthcare, Humanities & Social Sciences, and Engineering. Compared to prior benchmarks, MMVU features three key advancements. First, it challenges models to apply domain-specific knowledge and perform expert-level reasoning to analyze specialized-domain videos, moving beyond the basic visual perception typically assessed in current video benchmarks. Second, each example is annotated by human experts from scratch. We implement strict data quality controls to ensure the high quality of the dataset. Finally, each example is enriched with expert-annotated reasoning rationals and relevant domain knowledge, facilitating in-depth analysis. We conduct an extensive evaluation of 32 frontier multimodal foundation models on MMVU. The latest System-2-capable models, o1 and Gemini 2.0 Flash Thinking, achieve the highest performance among the tested models. However, they still fall short of matching human expertise. Through in-depth error analyses and case studies, we offer actionable insights for future advancements in expert-level, knowledge-intensive video understanding for specialized domains. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2501_12380 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | MMVU: Measuring Expert-Level Multi-Discipline Video Understanding Zhao, Yilun Xie, Lujing Zhang, Haowei Gan, Guo Long, Yitao Hu, Zhiyuan Hu, Tongyan Chen, Weiyuan Li, Chuhan Song, Junyang Xu, Zhijian Wang, Chengye Pan, Weifeng Shangguan, Ziyao Tang, Xiangru Liang, Zhenwen Liu, Yixin Zhao, Chen Cohan, Arman Computer Vision and Pattern Recognition Artificial Intelligence Computation and Language We introduce MMVU, a comprehensive expert-level, multi-discipline benchmark for evaluating foundation models in video understanding. MMVU includes 3,000 expert-annotated questions spanning 27 subjects across four core disciplines: Science, Healthcare, Humanities & Social Sciences, and Engineering. Compared to prior benchmarks, MMVU features three key advancements. First, it challenges models to apply domain-specific knowledge and perform expert-level reasoning to analyze specialized-domain videos, moving beyond the basic visual perception typically assessed in current video benchmarks. Second, each example is annotated by human experts from scratch. We implement strict data quality controls to ensure the high quality of the dataset. Finally, each example is enriched with expert-annotated reasoning rationals and relevant domain knowledge, facilitating in-depth analysis. We conduct an extensive evaluation of 32 frontier multimodal foundation models on MMVU. The latest System-2-capable models, o1 and Gemini 2.0 Flash Thinking, achieve the highest performance among the tested models. However, they still fall short of matching human expertise. Through in-depth error analyses and case studies, we offer actionable insights for future advancements in expert-level, knowledge-intensive video understanding for specialized domains. |
| title | MMVU: Measuring Expert-Level Multi-Discipline Video Understanding |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence Computation and Language |
| url | https://arxiv.org/abs/2501.12380 |