Audiovisual Matching on AVHBench
69.68AccuracyQwen 2.5 Omni + MoD-DPO++
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| Qwen 2.5 Omni + MoD-DPO++Reference Model=Qwen 2.5 Omni, Optimization Method=MoD-DPO++2026.03 | 69.68 | 43.32 | 96.04 | 59.71 | |
| Qwen 2.5 Omni + MoD-DPOReference Model=Qwen 2.5 Omni, Optimization Method=MoD-DPO2026.03 | 69.07 | 42.09 | 96.04 | 58.53 | |
| Qwen 2.5 Omni + OmniDPOReference Model=Qwen 2.5 Omni, Optimization Method=OmniDPO2026.03 | 61.5 | 23.9 | 99.14 | 38.52 | |
| MiniCPM-O 2.6 + MoD-DPO++Reference Model=MiniCPM-O 2.6, Optimization Method=MoD-DPO++2026.03 | 60.66 | 61.87 | 59.45 | 60.64 | |
| MiniCPM-O 2.6 + MoD-DPOReference Model=MiniCPM-O 2.6, Optimization Method=MoD-DPO2026.03 | 60.57 | 62.16 | 58.98 | 60.53 | |
| Qwen 2.5 Omni + DPOReference Model=Qwen 2.5 Omni, Optimization Method=DPO2026.03 | 59.32 | 19.76 | 98.93 | 32.94 | |
| Qwen 3 OmniReference Model=Qwen 3 Omni2026.03 | 58.52 | 97.76 | 19.14 | 32.02 | |
| MiniCPM-O 2.6 + OmniDPOReference Model=MiniCPM-O 2.6, Optimization Method=OmniDPO2026.03 | 56.86 | 57.82 | 55.9 | 56.84 | |
| Qwen 2.5 OmniReference Model=Qwen 2.5 Omni2026.03 | 54.69 | 9.81 | 99.57 | 17.85 | |
| MiniCPM-O 2.6 + DPOReference Model=MiniCPM-O 2.6, Optimization Method=DPO2026.03 | 54.56 | 55.91 | 53.2 | 54.52 | |
| OmniVinciReference Model=OmniVinci2026.03 | 54.32 | 87.42 | 21.21 | 34.14 | |
| MiniCPM-O 2.6Reference Model=MiniCPM-O 2.62026.03 | 54.26 | 56.72 | 51.81 | 54.15 | |
| VideoLLaMA 2Reference Model=VideoLLaMA 22026.03 | 52.93 | 92.11 | 13.75 | 23.93 | |
| VITA-1.5Reference Model=VITA-1.52026.03 | 46.85 | 66.31 | 27.4 | 38.78 |