ScaleLong: A Multi-Timescale Benchmark for Long Video Understanding
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_ | 1866918038906863616 |
|---|---|
| author | Ma, David Yuan, Huaqing Wang, Xingjian Zang, Qianbo Liu, Tianci He, Xinyang Wei, Yanbin Guo, Jiawei Jiahui, Ni Yang, Zhenzhu Cao, Meng Quan, Shanghaoran Li, Yizhi Zhou, Wangchunshu Liu, Jiaheng Huang, Wenhao Zhang, Ge Ni, Shiwen Jin, Xiaojie |
| author_facet | Ma, David Yuan, Huaqing Wang, Xingjian Zang, Qianbo Liu, Tianci He, Xinyang Wei, Yanbin Guo, Jiawei Jiahui, Ni Yang, Zhenzhu Cao, Meng Quan, Shanghaoran Li, Yizhi Zhou, Wangchunshu Liu, Jiaheng Huang, Wenhao Zhang, Ge Ni, Shiwen Jin, Xiaojie |
| contents | Although long-video understanding demands that models capture hierarchical temporal information -- from clip (seconds) and shot (tens of seconds) to event (minutes) and story (hours) -- existing benchmarks either neglect this multi-scale design or scatter scale-specific questions across different videos, preventing direct comparison of model performance across timescales on the same content. To address this, we introduce ScaleLong, the first benchmark to disentangle these factors by embedding questions targeting four hierarchical timescales -- clip (seconds), shot (tens of seconds), event (minutes), and story (hours) -- all within the same video content. This within-content multi-timescale questioning design enables direct comparison of model performance across timescales on identical videos. ScaleLong features 269 long videos (avg.\ 86\,min) from 5 main categories and 36 sub-categories, with 4--8 carefully designed questions, including at least one question for each timescale. Evaluating 23 MLLMs reveals a U-shaped performance curve, with higher accuracy at the shortest and longest timescales and a dip at intermediate levels. Furthermore, ablation studies show that increased visual token capacity consistently enhances reasoning across all timescales. ScaleLong offers a fine-grained, multi-timescale benchmark for advancing MLLM capabilities in long-video understanding. The code and dataset are available https://github.com/multimodal-art-projection/ScaleLong. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2505_23922 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | ScaleLong: A Multi-Timescale Benchmark for Long Video Understanding Ma, David Yuan, Huaqing Wang, Xingjian Zang, Qianbo Liu, Tianci He, Xinyang Wei, Yanbin Guo, Jiawei Jiahui, Ni Yang, Zhenzhu Cao, Meng Quan, Shanghaoran Li, Yizhi Zhou, Wangchunshu Liu, Jiaheng Huang, Wenhao Zhang, Ge Ni, Shiwen Jin, Xiaojie Computer Vision and Pattern Recognition Computation and Language Although long-video understanding demands that models capture hierarchical temporal information -- from clip (seconds) and shot (tens of seconds) to event (minutes) and story (hours) -- existing benchmarks either neglect this multi-scale design or scatter scale-specific questions across different videos, preventing direct comparison of model performance across timescales on the same content. To address this, we introduce ScaleLong, the first benchmark to disentangle these factors by embedding questions targeting four hierarchical timescales -- clip (seconds), shot (tens of seconds), event (minutes), and story (hours) -- all within the same video content. This within-content multi-timescale questioning design enables direct comparison of model performance across timescales on identical videos. ScaleLong features 269 long videos (avg.\ 86\,min) from 5 main categories and 36 sub-categories, with 4--8 carefully designed questions, including at least one question for each timescale. Evaluating 23 MLLMs reveals a U-shaped performance curve, with higher accuracy at the shortest and longest timescales and a dip at intermediate levels. Furthermore, ablation studies show that increased visual token capacity consistently enhances reasoning across all timescales. ScaleLong offers a fine-grained, multi-timescale benchmark for advancing MLLM capabilities in long-video understanding. The code and dataset are available https://github.com/multimodal-art-projection/ScaleLong. |
| title | ScaleLong: A Multi-Timescale Benchmark for Long Video Understanding |
| topic | Computer Vision and Pattern Recognition Computation and Language |
| url | https://arxiv.org/abs/2505.23922 |