AI-Generated Video Identification on GenVidBench
97.38MuseV ScoreVidGuard-R1
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| VidGuard-R1Type=MLLM, Optimization=GRPO2025.10 | 97.38 | 94.98 | 99.9 | 99.99 | 95.46 | 97.53 | |
| VidGuard-R1Type=MLLM, Optimization=GRPO, Pre-training=GenVideo-pretrained, Mode=Zero-shot2025.10 | 97.24 | 96.59 | 99.88 | 99.93 | 88.14 | 96.37 | |
| X3DType=CNN2025.10 | 92.39 | 37.27 | 65.72 | 49.6 | 97.51 | 77.09 | |
| CFV2Type=CNN2025.10 | 86.26 | 86.53 | 10.1 | 16.9 | 88.4 | 60.53 | |
| MViT V2Type=TF2025.10 | 76.34 | 98.29 | 47.5 | 96.62 | 97.58 | 79.9 | |
| TimeSformerType=TF2025.10 | 73.14 | 20.17 | 74.8 | 39.4 | 92.32 | 64.28 | |
| TSMType=CNN2025.10 | 70.37 | 54.7 | 78.46 | 70.37 | 96.76 | 76.4 | |
| VideoSwinType=TF2025.10 | 62.29 | 8.01 | 91.82 | 45.83 | 99.29 | 67.27 | |
| TRNType=CNN2025.10 | 38.92 | 26.64 | 91.34 | 93.98 | 93.97 | 71.26 | |
| TPNType=CNN2025.10 | 37.86 | 8.79 | 68.25 | 90.04 | 97.34 | 61.52 | |
| F3NetType=CNN2025.10 | 37.43 | 37.27 | 36.46 | 39.59 | 52.76 | 42.52 | |
| VidGuard-R1Type=MLLM, Mode=CoT2025.10 | 36.52 | 16.02 | 99.35 | 76.94 | 99.94 | 66.09 | |
| TINType=CNN2025.10 | 33.78 | 21.47 | 81.59 | 79.44 | 97.88 | 63.97 | |
| GPT-4.1 miniType=MLLM2025.10 | 26.07 | 33.78 | 94.07 | 57.19 | 87.64 | 59.62 | |
| Qwen2.5-VL-7BType=MLLM2025.10 | 25.86 | 27.06 | 68.51 | 43.26 | 71.15 | 47.3 | |
| UniFormer V2Type=TF2025.10 | 20.05 | 14.81 | 45.21 | 99.21 | 96.89 | 57.55 | |
| SlowFastType=CNN2025.10 | 12.25 | 12.68 | 38.34 | 45.93 | 93.63 | 41.66 | |
| I3DType=CNN2025.10 | 8.15 | 8.29 | 60.11 | 59.24 | 93.99 | 49.23 |