AI-generated video detection on GenVideo Show1
95.1ACCCAM-VFD
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| CAM-VFDTraining=10K Kinetics-400 + 10K Pika2026.05 | 95.1 | 90.5 | 92.8 | 97.6 | |
| NSG-VDTraining=10K Kinetics-400 + 10K Pika2026.05 | 90.83 | 87.5 | 90.52 | 96.38 | |
| NSG-VDTraining Data=Kinetics-400 and Pika, Evaluation Protocol=Standard2025.10 | 90.83 | 87.5 | 90.52 | 96.38 | |
| DeMambaTraining Data=Kinetics-400 and SEINE2025.10 | 86.6 | 75 | 84.84 | 96.89 | |
| NSG-VDTraining Data=Kinetics-400 and SEINE2025.10 | 86.25 | 100 | 87.91 | 96.69 | |
| Skyra-RL-GenVideoModel Size=7B2025.12 | 83.5 | 72 | 81.36 | — | |
| TALLTraining Data=Kinetics-400 and SEINE2025.10 | 80.5 | 62 | 76.07 | 96.45 | |
| TALLTraining=10K Kinetics-400 + 10K Pika2026.05 | 80.3 | 61.6 | 75.77 | 97.36 | |
| TALLTraining Data=Kinetics-400 and Pika, Evaluation Protocol=Standard2025.10 | 80.3 | 61.6 | 75.77 | 97.36 | |
| NPRTraining Data=Kinetics-400 and SEINE2025.10 | 76.2 | 56 | 70.18 | 89.96 | |
| DeMambaTraining=10K Kinetics-400 + 10K Pika2026.05 | 72.4 | 48.4 | 63.68 | 90.12 | |
| DeMambaTraining Data=Kinetics-400 and Pika, Evaluation Protocol=Standard2025.10 | 72.4 | 48.4 | 63.68 | 90.12 | |
| NPRTraining=10K Kinetics-400 + 10K Pika2026.05 | 65.7 | 33 | 49.03 | 90.5 | |
| NPRTraining Data=Kinetics-400 and Pika, Evaluation Protocol=Standard2025.10 | 65.7 | 33 | 49.03 | 90.5 | |
| DeMamba2025.12 | 61.5 | 85.57 | 71.61 | — | |
| STILTraining Data=Kinetics-400 and SEINE2025.10 | 59.3 | 18.8 | 31.6 | 92.79 | |
| Skyra-RLModel Size=7B2025.12 | 56.71 | 16.57 | 27.68 | — | |
| Skyra-SFTModel Size=7B2025.12 | 51.64 | 4.86 | 9.13 | — | |
| STILTraining=10K Kinetics-400 + 10K Pika2026.05 | 51 | 2 | 3.92 | 90.56 | |
| STILTraining Data=Kinetics-400 and Pika, Evaluation Protocol=Standard2025.10 | 51 | 2 | 3.92 | 90.56 | |
| NSG-VD2025.12 | 50.07 | 99.43 | 66.57 | — | |
| AIGVDet2025.12 | 50 | 0 | 0 | — |