Skeleton Based Action Recognition On J Hmdb
Metrics
Accuracy (RGB+pose)
Accuracy (pose)
Results
Performance results of various models on this benchmark
Comparison Table
| Model Name | Accuracy (RGB+pose) | Accuracy (pose) |
|---|---|---|
| make-skeleton-based-action-recognition-model-1 | - | 77.2 |
| potion-pose-motion-representation-for-action | 90.4 | 67.9 |
| multi-region-two-stream-r-cnn-for-action | 71.1 | - |
| finding-action-tubes | 62.5 | - |
| rpan-an-end-to-end-recurrent-pose-attention-1 | 83.9 | - |
| simple-yet-efficient-real-time-pose-based | - | 65.5 |
| potion-pose-motion-representation-for-action | 85.5 | - |
| quo-vadis-action-recognition-a-new-model-and | 84.1 | - |
| pa3d-pose-action-3d-machine-for-video | 69.5 | - |
| star-net-action-recognition-using-spatio | 64.3 | - |
| pa3d-pose-action-3d-machine-for-video | 86.1 | - |
| chained-multi-stream-networks-exploiting-pose | 76.1 | 56.8 |