Video Question Answering on Env-QA
49.23Attribute AccuracyGSMT
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| GSMTbackbone=CLIP2026.01 | 49.23 | 61.1 | 46.66 | 58.83 | 44.03 | 52.73 | |
| GSMTbackbone=AIO2026.01 | 48.76 | 57.99 | 44.96 | 57.39 | 43.18 | 50.81 | |
| MISTbackbone=CLIP2026.01 | 44.05 | 58.13 | 42.54 | 56.83 | 40.32 | 48.97 | |
| MISTbackbone=AIO2026.01 | 43.63 | 55.17 | 40.99 | 55.44 | 39.54 | 47.19 | |
| ATP2026.01 | 42.87 | 53.49 | 38.35 | 55.25 | 38.65 | 45.43 | |
| AIO2026.01 | 41.78 | 52.98 | 37.57 | 55.16 | 38.5 | 44.86 | |
| STAGE2026.01 | 39.49 | 49.93 | 34.52 | 55.32 | 38.54 | 41.97 |