Multi-choice Video Question Answering on EgoSchema (subset)
68.6AccuracyTarsier-34B
Evaluation Results
| Method | Links | |
|---|---|---|
| Tarsier-34BMode=Zero-shot2024.06 | 68.6 | |
| LangRepoMode=Zero-shot2024.06 | 66.2 | |
| VideoTreeMode=Zero-shot2024.06 | 66.2 | |
| VideoChat2Mode=Zero-shot2024.06 | 63.6 | |
| VideoAgentMode=Zero-shot2024.06 | 60.2 | |
| LLoViMode=Zero-shot2024.06 | 57.6 | |
| Tarsier-7BMode=Zero-shot2024.06 | 56 |