Table 12 Complexity and efficiency analysis.
From: Learning spatio-temporal context for basketball action pose estimation with a multi-stream network
Model | Backbone (%) | Input size (%) | Params (M) | FLOPs (G) | FPS |
|---|---|---|---|---|---|
HRNet | ResNet-50 | (\(256 \times 256\)) | 40.5 | 65.3 | 35.2 |
MSTCFL | ResNet-50 | (\(256 \times 256\)) | 35.2 | 55.7 | 28.5 |
Ours | ResNet-50 | (\(256 \times 256\)) | 60.1 | 45.6 | 25.1 |
