Reasoning is All You Need for Video Generalization: A Counterfactual Benchmark with Sub-question Evaluation
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866910985422372864 |
|---|---|
| author | Zhou, Qiji Gong, Yifan Bao, Guangsheng Qiu, Hongjie Li, Jinqiang Zhu, Xiangrong Zhang, Huajian Zhang, Yue |
| author_facet | Zhou, Qiji Gong, Yifan Bao, Guangsheng Qiu, Hongjie Li, Jinqiang Zhu, Xiangrong Zhang, Huajian Zhang, Yue |
| contents | Counterfactual reasoning is crucial for robust video understanding but remains underexplored in existing multimodal benchmarks. In this paper, we introduce \textbf{COVER} (\textbf{\underline{CO}}unterfactual \textbf{\underline{V}}id\textbf{\underline{E}}o \textbf{\underline{R}}easoning), a multidimensional multimodal benchmark that systematically evaluates MLLMs across the abstract-concrete and perception-cognition dimensions. Beyond prior multimodal benchmarks, COVER decomposes complex queries into structured sub-questions, enabling fine-grained reasoning analysis. Experiments on commercial and open-source models reveal a strong correlation between sub-question accuracy and counterfactual reasoning performance, highlighting the role of structured inference in video understanding. Furthermore, our results suggest a key insight: enhancing the reasoning capability of models is essential for improving the robustness of video understanding. COVER establishes a new standard for assessing MLLMs' logical reasoning abilities in dynamic environments. Our work is available at https://github.com/gongyifan-hash/COVER-Benchmark. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2503_10691 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Reasoning is All You Need for Video Generalization: A Counterfactual Benchmark with Sub-question Evaluation Zhou, Qiji Gong, Yifan Bao, Guangsheng Qiu, Hongjie Li, Jinqiang Zhu, Xiangrong Zhang, Huajian Zhang, Yue Computer Vision and Pattern Recognition Counterfactual reasoning is crucial for robust video understanding but remains underexplored in existing multimodal benchmarks. In this paper, we introduce \textbf{COVER} (\textbf{\underline{CO}}unterfactual \textbf{\underline{V}}id\textbf{\underline{E}}o \textbf{\underline{R}}easoning), a multidimensional multimodal benchmark that systematically evaluates MLLMs across the abstract-concrete and perception-cognition dimensions. Beyond prior multimodal benchmarks, COVER decomposes complex queries into structured sub-questions, enabling fine-grained reasoning analysis. Experiments on commercial and open-source models reveal a strong correlation between sub-question accuracy and counterfactual reasoning performance, highlighting the role of structured inference in video understanding. Furthermore, our results suggest a key insight: enhancing the reasoning capability of models is essential for improving the robustness of video understanding. COVER establishes a new standard for assessing MLLMs' logical reasoning abilities in dynamic environments. Our work is available at https://github.com/gongyifan-hash/COVER-Benchmark. |
| title | Reasoning is All You Need for Video Generalization: A Counterfactual Benchmark with Sub-question Evaluation |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2503.10691 |