Fine-grained action recognition on FineGym101
49.51AccuracyRepresentation Learning via Global Temporal Alignment and Cycle-Consistency
Evaluation Results
| Method | Links | |
|---|---|---|
| Representation Learning via Global Temporal Alignment and Cycle-ConsistencyTraining=only bn2021.05 | 49.51 | |
| Representation Learning via Global Temporal Alignment and Cycle-ConsistencyTraining=scratch2021.05 | 45.79 | |
| TCCTraining=only bn2021.05 | 45.62 | |
| SaLTraining=only bn2021.05 | 42.68 | |
| TCCTraining=scratch2021.05 | 41.78 | |
| TCNTraining=only bn2021.05 | 41.75 | |
| SaLTraining=scratch2021.05 | 40.25 | |
| D3TW*Training=only bn2021.05 | 38.21 | |
| TCNTraining=scratch2021.05 | 36.52 | |
| SpeedNetTraining=only bn2021.05 | 34.38 | |
| D3TW*Training=scratch2021.05 | 32.1 | |
| SpeedNetTraining=scratch2021.05 | 30.4 |