Research Article
Continuous Gesture Sequences Recognition Based on Few-Shot Learning
Table 2
Comparison with other models.
| | Model | 5-way 1-shot accuracy |
| | Nearest neighbor | 75.81% | | Nearest neighbor (SE) | 77.45% | | Matching network [20] | 85.62% | | 80.95% | | Autoencoder (ours) | 89.73% |
|
|
