Open-o3-Video-v2:基於可驗證時空證據的搜尋、定位與推理
Open-o3-Video-v2: Seek, Ground, and Reason with Verifiable Spatio-Temporal Evidence
Open-o3-Video-v2 將影片推理建構為統一的智能體式證據導向流程,主動補充觀察並把支持推理的時刻與物件證據寫入軌跡。在 V-STaR 上,Open-o3-Video-v2-4B 的 mAM 和 mLGM 分別為 40.4% 和 56.3%,較 Qwen3-VL-4B 高 8.5% 和 12.6%。程式碼及模型已釋出。
Published on Oct 13
Authors:
,
,
,
,
,
,
,
,
Abstract
Reliable video reasoning requires models to produce correct answers and expose supporting visual evidence. Open-o3-Video enables verifiable reasoning by embedding timestamps and object-level bounding boxes into reasoning traces, but is restricted to frames from an initial sparse sampler. We introduce Open-o3-Video-v2, which formulates video reasoning as a unified agentic evidencegrounded process. The model identifies missing information, acquires additional observations, grounds the moments and objects supporting its reasoning, and writes this evidence into the trajectory before continuing or answering. To support this behavior, we construct STGR-v2-CoT135K, including 52K multi-turn agentic grounded reasoning trajectories, and STGR-v2-RL-13K for reinforcement learning. We further propose Clue-Seek GRPO (CS-GRPO), which tracks how tool calls change accumulated evidence, rewards clue coverage and step-wise gains, and penalizes excessive, repetitive, or uninformative retrieval. On V-STaR, Open-o3-Video-v2-4B achieves 40.4% mAM and 56.3% mLGM, outperforming its Qwen3-VL-4B backbone by 8.5% and 12.6%, respectively. It also delivers consistent gains across general video understanding, long-video reasoning, and temporal grounding benchmarks. These results show that active evidence acquisition and verifiable object-level grounding are complementary for reliable video reasoning. The code and models are released at https://msalab-pku.github.io/projects/Open-o3-Video-v2.
View arXiv page View PDF Project page GitHub Add to collection
Get this paper in your agent:
hf papers read 2610.17656
Don't have the latest CLI?
curl -LsSf https://hf.co/cli/install.sh | bash
Models citing this paper 0
No model linking this paper
Cite arxiv.org/abs/2610.17656 in a model README.md to link it from this page.
Datasets citing this paper 0
No dataset linking this paper
Cite arxiv.org/abs/2610.17656 in a dataset README.md to link it from this page.
Spaces citing this paper 0
No Space linking this paper
Cite arxiv.org/abs/2610.17656 in a Space README.md to link it from this page.
Collections including this paper 0
No Collection including this paper
Add this paper to a collection to link it from this page.
來源:HuggingFace Daily Papers(社區熱門論文) · huggingface.co