VideoExplorer: Think With Videos For Agentic Long-Video Understanding
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866912682233298944 |
|---|---|
| author | Yuan, Huaying Liu, Zheng Zhou, Junjie Qian, Hongjin Shu, Yan Sebe, Nicu Wen, Ji-Rong Dou, Zhicheng |
| author_facet | Yuan, Huaying Liu, Zheng Zhou, Junjie Qian, Hongjin Shu, Yan Sebe, Nicu Wen, Ji-Rong Dou, Zhicheng |
| contents | Long-video understanding~(LVU) is a challenging problem in computer vision. Existing methods either downsample frames for single-pass reasoning, sacrificing fine-grained details, or depend on textual reasoning over task-agnostic representations, hindering task-specific perception and exploration. In this paper, we propose VideoExplorer, a framework grounded in the principle of ``thinking with video'', which naturally intertwines planning, temporal grounding, and scalable perception into a coherent reasoning process. Rather than reasoning over a static context, VideoExplorer iteratively formulates sub-questions, locates relevant moments, and performs task-oriented, temporally scalable video understanding until reaching the final answer, enabling faithful, efficient, and interpretable reasoning. To address the lack of LVU training resources, we construct a long-video reasoning dataset using difficulty-adaptive sampling to ensure high-quality trajectories on complex tasks. Building on this dataset, we design a two-stage training pipeline: supervised trajectory initialization followed by trajectory-level preference optimization, encouraging adaptive temporal grounding and iterative information integration guided by downstream rewards. Extensive evaluations on popular long-video understanding and reasoning benchmarks demonstrate VideoExplorer's significant advantage over existing baselines, highlighting its robustness, adaptability, and efficiency. Our code is made publicly available in this repository(https://github.com/yhy-2000/VideoDeepResearch). |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2506_10821 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | VideoExplorer: Think With Videos For Agentic Long-Video Understanding Yuan, Huaying Liu, Zheng Zhou, Junjie Qian, Hongjin Shu, Yan Sebe, Nicu Wen, Ji-Rong Dou, Zhicheng Computer Vision and Pattern Recognition Artificial Intelligence Computation and Language Long-video understanding~(LVU) is a challenging problem in computer vision. Existing methods either downsample frames for single-pass reasoning, sacrificing fine-grained details, or depend on textual reasoning over task-agnostic representations, hindering task-specific perception and exploration. In this paper, we propose VideoExplorer, a framework grounded in the principle of ``thinking with video'', which naturally intertwines planning, temporal grounding, and scalable perception into a coherent reasoning process. Rather than reasoning over a static context, VideoExplorer iteratively formulates sub-questions, locates relevant moments, and performs task-oriented, temporally scalable video understanding until reaching the final answer, enabling faithful, efficient, and interpretable reasoning. To address the lack of LVU training resources, we construct a long-video reasoning dataset using difficulty-adaptive sampling to ensure high-quality trajectories on complex tasks. Building on this dataset, we design a two-stage training pipeline: supervised trajectory initialization followed by trajectory-level preference optimization, encouraging adaptive temporal grounding and iterative information integration guided by downstream rewards. Extensive evaluations on popular long-video understanding and reasoning benchmarks demonstrate VideoExplorer's significant advantage over existing baselines, highlighting its robustness, adaptability, and efficiency. Our code is made publicly available in this repository(https://github.com/yhy-2000/VideoDeepResearch). |
| title | VideoExplorer: Think With Videos For Agentic Long-Video Understanding |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence Computation and Language |
| url | https://arxiv.org/abs/2506.10821 |