Image-based Audio-Visual Segmentation Quality Assessment on MQ-RAVSBench (Unseen) benchmark leaderboard · SOTA2 Research

Image-based Audio-Visual Segmentation Quality Assessment on MQ-RAVSBench (Unseen)

0.192RMSE (Perfect)Video-LLaMA3-7B

Evaluation Results

Method	Links
Video-LLaMA3-7B2026.02		0.192	0.259	0.306	0.264	0.298	0.254	0.289	0.524	0.951	0.371	88.28	82	0	0	0	0	0	0	16	9.92	94.09	0	0	0	0	0	0	0	16	10.47
MQ-Auditor2026.02		0.274	0.161	0.135	0.203	0.154	0.257	0.269	0.31	0.508	0.252	64.34	9,329	97.07	76.22	89.43	27.3	41.26	77.73	7,092	70.84	64.34	47.76	76.82	37.05	81.2	28.77	30.6	44.57	7,441	53.95
Gemini-3-Flash-Preview2026.02		0.409	0.357	0.334	0.32	0.298	0.345	0.314	0.301	0.25	0.325	18.39	7,807	79.4	0.86	3.16	3.14	3.97	73.84	8,042	37.92	21.98	19.78	64.31	23.29	37.47	24.85	47.15	69.08	8,849	44.04
Qwen2.5-Omni-7B2026.02		0.418	0.501	0.455	0.431	0.412	0.394	0.37	0.501	0.916	0.489	74.74	2,471	26.19	0	0	0	0	1.13	371	14.5	74.44	25.08	0	14.81	0	13.64	0	6.94	371	15.4
Ming-Flash-Omni2026.02		0.981	0.886	0.843	0.85	0.757	0.86	0.761	0.653	0.518	0.79	5.52	1,585	14.6	0	0	0.43	1.71	26.63	1,764	9.15	5.94	1.36	11.34	2.99	9.69	2.99	8.86	57.88	6,620	18.58