Action Recognition In Videos On Jester 1
評価指標
Val
評価結果
このベンチマークにおける各モデルのパフォーマンス結果
モデル名 | Val | Paper Title | Repository |
|---|---|---|---|
| DIN | 95.31 | DenseImage Network: Video Spatial-Temporal Evolution Encoding and Understanding | - |
| STM (Resnet-50, 16 frames) | 96.7 | STM: SpatioTemporal and Motion Encoding for Action Recognition | - |
| 3D-ShuffleNetV2 0.25x | 86.91 | Resource Efficient 3D Convolutional Neural Networks | |
| MultiScale TRN | 95.31 | Temporal Relational Reasoning in Videos | |
| convSTAR | 92.7 | Gating Revisited: Deep Multi-layer RNNs That Can Be Trained | |
| CPNet Res34, 5 CP | 96.7 | Learning Video Representations from Correspondence Proposals | |
| MFNet | 96.68 | Motion Feature Network: Fixed Motion Filter for Action Recognition | - |
| 3D-SqueezeNet | 90.77 | Resource Efficient 3D Convolutional Neural Networks | |
| 3D-MobileNetV2 0.2x | 86.43 | Resource Efficient 3D Convolutional Neural Networks |
0 of 9 row(s) selected.