Hallucination Examination on VidHalluc
78.08BQASEASON
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| SEASONBackbone=QWEN2.5-VL-7B, Training-free=true2025.12 | 78.08 | 87.32 | 71.92 | 77.67 | 78.75 | |
| RRPOBackbone=LLaVA-Video-7B, Training-free=false2025.12 | 76.8 | 91.23 | 49.83 | 37.67 | 63.88 | |
| ArrowRLBackbone=QWEN2.5-VL-7B, Training-free=false2025.12 | 76.14 | 87.84 | 70.92 | 57.83 | 73.18 | |
| DINO-HEALBackbone=QWEN2.5-VL-7B, Training-free=true2025.12 | 75.86 | 84.21 | 75.86 | 58.67 | 73.65 | |
| QWEN2.5-VL-7BBackbone=N/A, Training-free=-2025.12 | 75.79 | 84.05 | 74.91 | 59 | 73.44 | |
| DINO-HEALBackbone=LLaVA-Video-7B, Training-free=true2025.12 | 75.27 | 90.81 | 51.51 | 37.67 | 63.81 | |
| LLaVA-Video-7BBackbone=N/A, Training-free=-2025.12 | 75.02 | 90.76 | 51.23 | 38 | 63.75 | |
| TPOBackbone=LLaVA-Video-7B, Training-free=false2025.12 | 74.85 | 90.69 | 49.62 | 42.5 | 64.42 | |
| SEASONBackbone=LLaVA-Video-7B, Training-free=true2025.12 | 74.71 | 90.95 | 49.86 | 50.33 | 66.46 | |
| MultiToPBackbone=Qwen3-VL-4B-Instruct2026.06 | 74.64 | 89.08 | 61.35 | 86.83 | — | |
| TCDBackbone=QWEN2.5-VL-7B, Training-free=true2025.12 | 74.6 | 85.57 | 74.37 | 64.67 | 74.8 | |
| LLaVA-OV-7BBackbone=N/A, Training-free=-2025.12 | 74.36 | 90.27 | 63.65 | 53 | 70.32 | |
| DINO-HEALBackbone=LLaVA-OV-7B, Training-free=true2025.12 | 74.29 | 90.36 | 63.18 | 53 | 70.21 | |
| TCDBackbone=LLaVA-Video-7B, Training-free=true2025.12 | 73.5 | 90.41 | 49.64 | 46 | 64.89 | |
| SEASONBackbone=LLaVA-OV-7B, Training-free=true2025.12 | 73.15 | 90.51 | 60.29 | 77.5 | 75.36 | |
| TCDBackbone=LLaVA-OV-7B, Training-free=true2025.12 | 72.44 | 90.06 | 58.57 | 64.33 | 71.35 | |
| TA-AEBackbone=Qwen3-VL-4B-Instruct2026.06 | 61.39 | 86.53 | 41.67 | 60 | — | |
| DINO-HEALBackbone=Qwen3-VL-4B-Instruct2026.06 | 61.08 | 86.9 | 38.1 | 58.3 | — | |
| MotionCDBackbone=Qwen3-VL-4B-Instruct2026.06 | 60.54 | 84.57 | 35.34 | 57.16 | — | |
| TCDBackbone=Qwen3-VL-4B-Instruct2026.06 | 59.82 | 85.31 | 37.17 | 55 | — | |
| SmartSightBackbone=Qwen3-VL-4B-Instruct2026.06 | 59.05 | 85.95 | 40.1 | 58.84 | — | |
| Qwen3-VL-4B-InstructBackbone=Qwen3-VL-4B-Instruct2026.06 | 58.57 | 86.32 | 37.3 | 56.5 | — | |
| MultiToPBackbone=Video-LLaVA-7B2026.06 | 25.08 | 65.22 | 42.02 | 49.33 | — | |
| DINO-HEALBackbone=Video-LLaVA-7B2026.06 | 24.96 | 65.76 | 36.57 | 29.24 | — | |
| TCDBackbone=Video-LLaVA-7B2026.06 | 24.81 | 64.56 | 30.08 | 26.06 | — | |
| MotionCDBackbone=Video-LLaVA-7B2026.06 | 24.74 | 64 | 28.6 | 25.92 | — | |
| Video-LLaVA-7BBackbone=Video-LLaVA-7B2026.06 | 23.88 | 65.18 | 30.12 | 28.83 | — | |
| TA-AEBackbone=Video-LLaVA-7B2026.06 | 23.86 | 65.41 | 38.9 | 30.48 | — | |
| SmartSightBackbone=Video-LLaVA-7B2026.06 | 23.72 | 65.04 | 37.64 | 30.3 | — |