Activity Recognition On Rwf 2000
Metrics
Accuracy
Results
Performance results of various models on this benchmark
Model Name | Accuracy | Paper Title | Repository |
|---|---|---|---|
| Structured Keypoint Pooling | 93.4 | Unified Keypoint-based Action Recognition Framework via Structured Keypoint Pooling | - |
| Human Skeletons + Change Detection | 90.25 | Human skeletons and change detection for efficient violence detection in surveillance videos | |
| Separable Convolutional LSTM | 89.75 | Efficient Two-Stream Network for Violence Detection Using Separable Convolutional LSTM | |
| SPIL Convolution | 89.3 | Human Interaction Learning on 3D Skeleton Point Clouds for Video Violence Recognition | - |
| Flow Gated Network | 87.25 | RWF-2000: An Open Large Scale Video Database for Violence Detection | |
| Semi-Supervised Hard Attention (SSHA); pretrained on Deepmind Kinetics dataset | 90.4 | Video Violence Recognition and Localization Using a Semi-Supervised Hard Attention Model | - |
0 of 6 row(s) selected.