https://github.com/marinero4972/Open-o3-Video\n","updatedAt":"2025-10-24T03:12:20.653Z","author":{"_id":"65a28e129acab19980226731","avatarUrl":"/avatars/abc3828f807efc4e03837b0eae063f98.svg","fullname":"Jiahao Meng","name":"marinero4972","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":2,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"en","probability":0.5923375487327576},"editors":["marinero4972"],"editorAvatarUrls":["/avatars/abc3828f807efc4e03837b0eae063f98.svg"],"reactions":[{"reaction":"🔥","users":["taesiri","tyfeld","LXT"],"count":3},{"reaction":"❤️","users":["tyfeld","LXT"],"count":2},{"reaction":"🚀","users":["ldkong"],"count":1}],"isReport":false}},{"id":"68fc29a61d895e384705eac8","author":{"_id":"63d3e0e8ff1384ce6c5dd17d","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/1674830754237-63d3e0e8ff1384ce6c5dd17d.jpeg","fullname":"Librarian Bot (Bot)","name":"librarian-bot","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":318,"isUserFollowing":false},"createdAt":"2025-10-25T01:36:38.000Z","type":"comment","data":{"edited":false,"hidden":false,"latest":{"raw":"This is an automated message from the [Librarian Bot](https://huggingface.co/librarian-bots). I found the following papers similar to this paper. \n\nThe following papers were recommended by the Semantic Scholar API \n\n* [Video-LMM Post-Training: A Deep Dive into Video Reasoning with Large Multimodal Models](https://huggingface.co/papers/2510.05034) (2025)\n* [Conan: Progressive Learning to Reason Like a Detective over Multi-Scale Visual Evidence](https://huggingface.co/papers/2510.20470) (2025)\n* [VideoChat-R1.5: Visual Test-Time Scaling to Reinforce Multimodal Reasoning by Iterative Perception](https://huggingface.co/papers/2509.21100) (2025)\n* [Video-STR: Reinforcing MLLMs in Video Spatio-Temporal Reasoning with Relation Graph](https://huggingface.co/papers/2510.10976) (2025)\n* [Video-MTR: Reinforced Multi-Turn Reasoning for Long Video Understanding](https://huggingface.co/papers/2508.20478) (2025)\n* [When Thinking Drifts: Evidential Grounding for Robust Video Reasoning](https://huggingface.co/papers/2510.06077) (2025)\n* [Select Less, Reason More: Prioritizing Evidence Purity for Video Reasoning](https://huggingface.co/papers/2510.15440) (2025)\n\n\n Please give a thumbs up to this comment if you found it helpful!\n\n If you want recommendations for any Paper on Hugging Face checkout [this](https://huggingface.co/spaces/librarian-bots/recommend_similar_papers) Space\n\n You can directly ask Librarian Bot for paper recommendations by tagging it in a comment: `@librarian-bot recommend`","html":"
This is an automated message from the Librarian Bot. I found the following papers similar to this paper.
\nThe following papers were recommended by the Semantic Scholar API
\n- \n
- Video-LMM Post-Training: A Deep Dive into Video Reasoning with Large Multimodal Models (2025) \n
- Conan: Progressive Learning to Reason Like a Detective over Multi-Scale Visual Evidence (2025) \n
- VideoChat-R1.5: Visual Test-Time Scaling to Reinforce Multimodal Reasoning by Iterative Perception (2025) \n
- Video-STR: Reinforcing MLLMs in Video Spatio-Temporal Reasoning with Relation Graph (2025) \n
- Video-MTR: Reinforced Multi-Turn Reasoning for Long Video Understanding (2025) \n
- When Thinking Drifts: Evidential Grounding for Robust Video Reasoning (2025) \n
- Select Less, Reason More: Prioritizing Evidence Purity for Video Reasoning (2025) \n
Please give a thumbs up to this comment if you found it helpful!
\nIf you want recommendations for any Paper on Hugging Face checkout this Space
\n You can directly ask Librarian Bot for paper recommendations by tagging it in a comment: \n\n@librarian-bot\n\t recommend
Papers
arxiv:2510.20579
Open-o3 Video: Grounded Video Reasoning with Explicit Spatio-Temporal Evidence
Abstract
Open-o3 Video integrates spatio-temporal evidence into video reasoning, achieving state-of-the-art performance on multiple benchmarks and providing valuable reasoning traces for test-time scaling.
AI-generated summary
Most video reasoning models only generate textual reasoning traces without indicating when and where key evidence appears. Recent models such as OpenAI-o3 have sparked wide interest in evidence-centered reasoning for images, yet extending this ability to videos is more challenging, as it requires joint temporal tracking and spatial localization across dynamic scenes. We introduce Open-o3 Video, a non-agent framework that integrates explicit spatio-temporal evidence into video reasoning, and carefully collect training data and design training strategies to address the aforementioned challenges. The model highlights key timestamps, objects, and bounding boxes alongside its answers, allowing reasoning to be grounded in concrete visual observations. To enable this functionality, we first curate and build two high-quality datasets, STGR-CoT-30k for SFT and STGR-RL-36k for RL, with carefully constructed temporal and spatial annotations, since most existing datasets offer either temporal spans for videos or spatial boxes on images, lacking unified spatio-temporal supervision and reasoning traces. Then, we adopt a cold-start reinforcement learning strategy with multiple specially designed rewards that jointly encourage answer accuracy, temporal alignment, and spatial precision. On V-STAR benchmark, Open-o3 Video achieves state-of-the-art performance, raising mAM by 14.4% and mLGM by 24.2% on the Qwen2.5-VL baseline. Consistent improvements are also observed on a broad range of video understanding benchmarks, including VideoMME, WorldSense, VideoMMMU, and TVGBench. Beyond accuracy, the reasoning traces produced by Open-o3 Video also provide valuable signals for test-time scaling, enabling confidence-aware verification and improving answer reliability.
Community
Open-o3 Video introduces a video reasoning framework that grounds its answers in explicit spatio-temporal evidence—highlighting when and where key visual cues occur—achieving state-of-the-art performance across video understanding benchmarks.
This is an automated message from the Librarian Bot. I found the following papers similar to this paper.
The following papers were recommended by the Semantic Scholar API
- Video-LMM Post-Training: A Deep Dive into Video Reasoning with Large Multimodal Models (2025)
- Conan: Progressive Learning to Reason Like a Detective over Multi-Scale Visual Evidence (2025)
- VideoChat-R1.5: Visual Test-Time Scaling to Reinforce Multimodal Reasoning by Iterative Perception (2025)
- Video-STR: Reinforcing MLLMs in Video Spatio-Temporal Reasoning with Relation Graph (2025)
- Video-MTR: Reinforced Multi-Turn Reasoning for Long Video Understanding (2025)
- When Thinking Drifts: Evidential Grounding for Robust Video Reasoning (2025)
- Select Less, Reason More: Prioritizing Evidence Purity for Video Reasoning (2025)
Please give a thumbs up to this comment if you found it helpful!
If you want recommendations for any Paper on Hugging Face checkout this Space
You can directly ask Librarian Bot for paper recommendations by tagging it in a comment:
@librarian-bot
recommend
Models citing this paper 0
No model linking this paper
Cite arxiv.org/abs/2510.20579 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/2510.20579 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/2510.20579 in a Space README.md to link it from this page.