Research map / Video / Video-language understanding
CoT-RVS: Zero-Shot Chain-of-Thought Reasoning Segmentation for Videos
Shiu-hong Kao, Yu-Wing Tai, Chi-Keung Tang
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 |
|---|---|---|---|
| use zero-shot CoT-RVS with GPT-4o instead of best finetuned baseline | reasoning VOS · MeViS | J&F: 51.3 -> 52.2 | Improved |
| use CoT-RVS-GPT-4o instead of GLUS | ReasonVOS · ReasonVOS | J&F: 49.9 -> 65.5 | Improved |
| use CoT-RVS-GPT-4o instead of HyperSeg | ReVOS · ReVOS | J&F: 55.7 -> 55.9 | Improved |
| use CoT-RVS-Gemma3-12B instead of best finetuned baseline | MeViS · MeViS | J&F: 51.3 -> 44.2 | Hurt |
| use GPT-4o instead of Gemma3-12B as keyframe selector (ablation) | CoT-RVS · ReasonVOS | J&F: 50.7 -> 65.5 | Improved |
| use CoT-RVS-GPT-4o on temporally sensitive queries | CoT-RVS · T-ReasonVOS | J&F: 37.4 -> 55.5 | Improved |
| use 8 keyframe candidates instead of 16 (ablation) | CoT-RVS (GPT-4o) · ReasonVOS | J&F: 60.0 -> 65.5 | Improved |
| use 4 keyframe candidates instead of 8 (ablation) | CoT-RVS (GPT-4o) · ReasonVOS | J&F: 65.5 -> 66.2 | No clear effect |
| add chain-of-thought prompting for keyframe selection (ablation) | CoT-RVS · ReasonVOS / online | J&F | Improved |
| use online CoT-RVS (LLaVA1.5-7B) instead of offline GPT-4o | CoT-RVS · Refer-DAVIS-17 | J&F: 79.1 -> 73.9 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 3 of 3 on most shared results.
- GLUS — GLUS: Global-Local Reasoning Unified into A Single Large Language Model for Video Segmentation (CVPR 2025) · better on 3 of 3 shared results
- SAMWISE — SAMWISE: Infusing Wisdom in SAM2 for Text-Driven Video Segmentation (CVPR 2025) · better on 2 of 2 shared results
- VISA-13B — VISA: Reasoning Video Object Segmentation via Large Language Model (ECCV 2024) · better on 2 of 2 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
- The Devil is in Temporal Token: High Quality Video Reasoning Segmentation (CVPR 2025 · similarity 0.90)
- VideoReasonBench: Can MLLMs Perform Vision-Centric Complex Video Reasoning? (ICLR 2026 · similarity 0.89)
- One Token to Seg Them All: Language Instructed Reasoning Segmentation in Videos (NeurIPS 2024 · similarity 0.88)
- VideoEspresso: A Large-Scale Chain-of-Thought Dataset for Fine-Grained Video Reasoning via Core Frame Selection (CVPR 2025 · similarity 0.88)
- Fast Reasoning Segmentation for Images and Videos (CVPR 2026 · similarity 0.87)
- Visual CoT: Advancing Multi-Modal Language Models with a Comprehensive Dataset and Benchmark for Chain-of-Thought Reasoning (NeurIPS 2024 · 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.
