Audio-Visual Perception on Daily-Omni
60.65ScoreEchoingPixels
Evaluation results
| Method | Links | |
|---|---|---|
| EchoingPixelsModel Scale=3B, Token Budget=20%2025.12 | 60.65 | |
| Full Model (Qwen2.5-Omni-7B)Model Scale=7B, Token Budget=100%2025.12 | 60.48 | |
| Full Model (Qwen2.5-Omni-3B)Model Scale=3B, Token Budget=100%2025.12 | 59.65 | |
| EchoingPixelsModel Scale=3B, Token Budget=10%2025.12 | 57.56 | |
| EchoingPixelsModel Scale=7B, Token Budget=10%2025.12 | 56.98 | |
| EchoingPixelsModel Scale=3B, Token Budget=5%2025.12 | 52.88 | |
| IntraModalModel Scale=7B, Token Budget=25%2025.12 | 49.79 | |
| IntraModalModel Scale=3B, Token Budget=25%2025.12 | 46.2 |