Research map / Video / Video-language understanding
Ego-R1: Agentic Chain-of-Tool-Thought for Ultra-Long Egocentric Video Reasoning
Shulin Tian, Rui Wang, Hongming Guo, Penghao Wu, Yuhao Dong, Xiuying Wang, Jingkang Yang, Hao Zhang, Hongyuan Zhu, Ziwei Liu
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add SFT on CoTT data to the base model (ablation) | Qwen2.5-3B-Instruct · Ego-R1 Bench | accuracy (%): 1.4 -> 34.3 | Improved |
| add RL (GRPO) on top of SFT (ablation) | Ego-R1-SFT · Ego-R1 Bench | accuracy (%): 34.3 -> 46.0 | Improved |
| apply RL alone without CoTT SFT (ablation) | Qwen2.5-3B-Instruct · Ego-R1 Bench | accuracy (%): 1.4 -> 0.0 | Hurt |
| apply RL alone without CoTT SFT (ablation) | Qwen2.5-3B-Instruct · Ego-R1 Bench | format accuracy (%): 4.3 -> 13.3 | Improved |
| add SFT on CoTT data (ablation) | Qwen2.5-3B-Instruct · Ego-R1 Bench | format accuracy (%): 4.3 -> 100.0 | Improved |
| replace LLaVA-Video with Gemini-1.5-Pro as video_llm (ablation) | Ego-R1 · Ego-R1 Bench | accuracy (%): 43.7 -> 46.0 | Improved |
| add video_llm and vlm tools to RAG-only (ablation) | Ego-R1 · Ego-R1 Bench | accuracy (%): 39.7 -> 46.0 | Improved |
Later papers that compared with this method
2 later accepted papers report Ego-R1 as a baseline in their results tables; 2 of them report a better number on most of the results they share with it.
- VideoSEAL: Mitigating Evidence Misalignment in Agentic Long Video Understanding by Decoupling Answer Authority (ICML 2026 · better on 2 of 2 shared results)
- LongVideo-R1: Smart Navigation for Low-cost Long Video Understanding (CVPR 2026 · better on 1 of 1 shared result)
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- VideoAgent — VideoAgent: A Memory-augmented Multimodal Agent for Video Understanding (ECCV 2024) · better on 4 of 4 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- EgoThinker: Unveiling Egocentric Reasoning with Spatio-Temporal CoT (NeurIPS 2025 · similarity 0.91)
- Omnia de EgoTempo: Benchmarking Temporal Understanding of Multi-Modal LLMs in Egocentric Videos (CVPR 2025 · similarity 0.90)
- EgoExoBench: A Benchmark for First- and Third-person View Video Understanding in MLLMs (NeurIPS 2025 · similarity 0.90)
- Ego2Web: A Web Agent Benchmark Grounded in Egocentric Videos (CVPR 2026 · similarity 0.88)
- MMEgo: Towards Building Egocentric Multimodal LLMs for Video QA (ICLR 2025 · similarity 0.88)
- Video-R1: Reinforcing Video Reasoning in MLLMs (NeurIPS 2025 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
