Chain-of-Glimpse: Search-Guided Progressive Object-Grounded Reasoning for Video Understanding
Zhixuan Wu*, Quanxing Zha*, Teng Wang, Genbao Xu · 2026
Read Chain-of-Glimpse: Search-Guided Progressive Object-Grounded Reasoning for Video Understanding by Zhixuan Wu*, Quanxing Zha*, Teng Wang, Genbao Xu (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-14692.txt.
Description
Wu and eight colleagues built Chain-of-Glimpse, a video-reasoning method that ties each reasoning step to a specific visible object region instead of leaning on overall saliency. A search-guided controller, trained with reinforcement learning that rewards well-grounded answers, builds these traces step by step into interpretable multi-step decisions. Tests on NExTQA, Video-Holmes, CG-Bench Reasoning, and VRBench showed steady gains and stronger generalization across varied video tasks.
Catalog record
| CREATOR | Zhixuan Wu*, Quanxing Zha*, Teng Wang, Genbao Xu |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 43652 |
| LICENSE | http://creativecommons.org/publicdomain/zero/1.0/ |
| RIGHTS | CC0 |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2604.14692