Violence Detection on Movies Fight
100AccuracyFightNet
Evaluation results
| Method | Links | |
|---|---|---|
| FightNetType=Deep-Learning Based2019.11 | 100 | |
| ConvLSTMType=Deep-Learning Based2019.11 | 100 | |
| C3DType=Deep-Learning Based2019.11 | 100 | |
| I3D(RGB only)Type=Deep-Learning Based2019.11 | 100 | |
| I3D(Flow only)Type=Deep-Learning Based2019.11 | 100 | |
| I3D(Fusion)Type=Deep-Learning Based2019.11 | 100 | |
| Flow Gated NetworkType=Deep-Learning Based2019.11 | 100 | |
| ConvLSTM2022.02 | 100 | |
| I3DInput modality=RGB2022.02 | 100 | |
| I3DInput modality=Optical-flow2022.02 | 100 | |
| I3DInput modality=Two-stream2022.02 | 100 | |
| Cheng et al. ModelArchitecture=P3D2022.02 | 100 | |
| 3D ConvNetType=Deep-Learning Based2019.11 | 99.97 | |
| SSHAInput modality=RGB, Localization=Disabled2022.02 | 99 | |
| SSHAInput modality=RGB2022.02 | 99 | |
| SSHAInput modality=Two-stream2022.02 | 99 | |
| SSHAInput modality=Optical-flow2022.02 | 98.5 | |
| MOSIFT+HIKType=Hand-Crafted Features2019.11 | 89.5 | |
| HOF+HIKType=Hand-Crafted Features2019.11 | 59 | |
| HOG+HIKType=Hand-Crafted Features2019.11 | 49 |