Video Question Answering on EgoSchema 5031 videos (test)
62.4Top-1 AccuracyLifelongMemory
Evaluation Results
| Method | Links | |
|---|---|---|
| LifelongMemoryZero Shot=true, Video Training=false, Closed Model=true, Params=1.8T2024.03 | 62.4 | |
| TarsierZero Shot=true, Video Training=true, Closed Model=false, Params=34B2024.03 | 61.7 | |
| LVNet + MVUZero Shot=true, Video Training=false, Closed Model=true, Params=1.8T2024.03 | 61.3 | |
| VideoTreeZero Shot=true, Video Training=false, Closed Model=true, Params=1.8T2024.03 | 61.1 | |
| LVNetZero Shot=true, Video Training=false, Closed Model=true, Params=1.8T2024.03 | 61.1 | |
| InternVideo2Zero Shot=true, Video Training=true, Closed Model=false, Params=7B2024.03 | 55.8 | |
| VideoAgentZero Shot=true, Video Training=false, Closed Model=true, Params=1.8T2024.03 | 54.1 | |
| MoreVQAZero Shot=true, Video Training=false, Closed Model=false2024.03 | 51.7 | |
| LLoViZero Shot=true, Video Training=false, Closed Model=true, Params=1.8T2024.03 | 50.3 | |
| TarsierZero Shot=true, Video Training=true, Closed Model=false, Params=7B2024.03 | 49.9 | |
| VamosZero Shot=true, Video Training=false, Closed Model=true, Params=1.8T2024.03 | 48.3 | |
| MC-ViT-LZero Shot=false, Video Training=true, Closed Model=false, Params=424M2024.03 | 44.4 | |
| LangRepoZero Shot=true, Video Training=false, Closed Model=false, Params=12B2024.03 | 41.2 | |
| SF-VLM + MVUZero Shot=true, Video Training=false, Closed Model=false, Params=13B2024.03 | 37.6 | |
| VamosZero Shot=true, Video Training=false, Closed Model=false, Params=13B2024.03 | 36.7 | |
| SF-VLMZero Shot=true, Video Training=false, Closed Model=false, Params=13B2024.03 | 36.4 | |
| LLoViZero Shot=true, Video Training=false, Closed Model=false, Params=13B2024.03 | 33.5 | |
| LongViViTZero Shot=true, Video Training=true, Closed Model=false, Params=1B2024.03 | 33.3 | |
| InternVideoZero Shot=false, Video Training=true, Closed Model=false, Params=478M2024.03 | 32.1 | |
| SeViLA+ShortViViTZero Shot=false, Video Training=true, Closed Model=false, Params=5B2024.03 | 31.3 | |
| mPLUG-OwlZero Shot=true, Video Training=true, Closed Model=false, Params=7.2B2024.03 | 31.1 | |
| ImageViTZero Shot=false, Video Training=false, Closed Model=false, Params=1B2024.03 | 30.9 | |
| FrozenBiLMZero Shot=true, Video Training=true, Closed Model=false, Params=1.2B2024.03 | 26.9 | |
| SeViLAZero Shot=true, Video Training=true, Closed Model=false, Params=4B2024.03 | 22.7 | |
| Random Selection2024.03 | 20 | |
| VIOLETZero Shot=false, Video Training=true, Closed Model=false, Params=198M2024.03 | 19.9 |