Zero Shot Skeletal Action Recognition On Ntu
Metrics
Accuracy (12 unseen classes)
Accuracy (5 unseen classes)
Random Split Accuracy
Results
Performance results of various models on this benchmark
Comparison Table
| Model Name | Accuracy (12 unseen classes) | Accuracy (5 unseen classes) | Random Split Accuracy |
|---|---|---|---|
| sa-dvae-improving-zero-shot-skeleton-based | 41.38 | 82.37 | 84.20 |
| sa-dvae-improving-zero-shot-skeleton-based | - | - | 87.61 |
| part-aware-unified-representation-of-language-1 | 40.99 | 79.23 | - |
| fine-grained-side-information-guided-dual | 45.10 | 81.40 | 77.50 |
| tdsm-triplet-diffusion-for-skeleton-text | 56.03 | 86.49 | 88.88 |
| zero-shot-skeleton-based-action-recognition | 40.18 | 77.98 | 65.08 |
| generalized-zero-and-few-shot-learning-via | 28.96 | 76.84 | 60.74 |
| syntactically-guided-generative-embeddings-1 | 33.30 | 75.81 | 64.19 |