Referring and Reasoning Video Object Segmentation on ReVOS
66.3Overall J&F ScoreAgentRVOS
Evaluation Results
| Method | Links | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| AgentRVOSMLLM=GPT-5, Training Protocol=Training-free Methods2026.03 | 66.3 | 66.8 | 70.8 | 68.8 | 61.4 | 66 | 63.7 | 64.1 | 68.4 | |
| AgentRVOSMLLM=Qwen3-VL-32B-T, Training Protocol=Training-free Methods2026.03 | 62.5 | 62.8 | 66.8 | 64.8 | 58 | 62.6 | 60.3 | 60.4 | 64.7 | |
| VideoSeg-R1MLLM=Qwen2.5-VL-7B, Training Protocol=Training-based Methods2026.03 | 61.1 | — | — | — | — | — | — | 58.2 | 64 | |
| Sa2VAMLLM=InternVL3-14B, Training Protocol=Training-based Methods2026.03 | 60.7 | — | — | — | — | — | — | — | — | |
| VRS-HQMLLM=Chat-UniVi-13B, Training Protocol=Training-based Methods2026.03 | 60 | 61.1 | 65.5 | 63.3 | 54.1 | 59.4 | 56.8 | 57.6 | 62.5 | |
| AgentRVOSMLLM=Qwen3-VL-8B-T, Training Protocol=Training-free Methods2026.03 | 59.8 | 58.7 | 62.5 | 60.6 | 56.8 | 61.3 | 59 | 57.7 | 61.9 | |
| RGA3MLLM=Qwen2.5-VL-7B, Training Protocol=Training-based Methods2026.03 | 58 | 58.7 | 62.3 | 60.5 | 53.1 | 57.7 | 55.4 | 55.9 | 60 | |
| Sa2VAMLLM=InternVL2-8B, Training Protocol=Training-based Methods2026.03 | 57.6 | — | — | — | — | — | — | — | — | |
| ViLLaMLLM=InternVideo2-6B, Training Protocol=Training-based Methods2026.03 | 57 | — | — | — | — | — | — | 54.9 | 59.1 | |
| CoT-RVSMLLM=GPT-4o, Training Protocol=Training-free Methods2026.03 | 55.9 | — | — | — | — | — | — | 52.8 | 59 | |
| HyperSegMLLM=Mipha-3B, Training Protocol=Training-based Methods2026.03 | 55.7 | 56 | 60.9 | 58.5 | 50.2 | 55.8 | 53 | 53.1 | 58.4 | |
| GLUSMLLM=LISA-7B, Training Protocol=Training-based Methods2026.03 | 54.9 | 56 | 60.7 | 58.3 | 48.8 | 53.9 | 51.4 | 52.4 | 57.3 | |
| InstructSegMLLM=Mipha-3B, Training Protocol=Training-based Methods2026.03 | 54.5 | 54.8 | 59.2 | 57 | 49.2 | 54.7 | 51.9 | 52 | 56.9 | |
| CoT-RVSMLLM=Qwen3-VL-8B-T, Training Protocol=Training-free Methods, reproduced=true2026.03 | 54.3 | 56.1 | 61.5 | 58.8 | 46.6 | 53 | 49.8 | 51.4 | 57.3 | |
| VISAMLLM=Chat-UniVi-13B, Training Protocol=Training-based Methods2026.03 | 50.9 | 55.6 | 59.1 | 57.4 | 42 | 46.7 | 44.3 | 48.8 | 52.9 | |
| CoT-RVSMLLM=Gemma3-12B, Training Protocol=Training-free Methods2026.03 | 47.1 | — | — | — | — | — | — | 43.4 | 50.9 |