Sign Language Recognition On Rwth Phoenix
Métriques
Word Error Rate (WER)
Résultats
Résultats de performance de divers modèles sur ce benchmark
Tableau comparatif
| Nom du modèle | Word Error Rate (WER) |
|---|---|
| 007-democratically-finding-the-cause-of | 23.4 |
| multimodal-locally-enhanced-transformer-for | 20.89 |
| subunets-end-to-end-hand-shape-and-continuous | 40.7 |
| context-matters-self-attention-for-sign | 29.7 |
| visual-alignment-constraint-for-continuous | 22.1 |
| signbert-hand-model-aware-self-supervised-pre | 20 |
| a-deep-neural-framework-for-continuous-sign | 22.86 |
| connectionist-temporal-fusion-for-sign | 37.8 |
| c2slr-consistency-enhanced-continuous-sign | 20.4 |
| two-stream-network-for-sign-language | 18.4 |
| multi-stream-keypoint-attention-network-for | 22.1 |
| continuous-sign-language-recognition-with | 19.4 |
| stochastic-fine-grained-labeling-of-multi | 25.3 |
| slowfast-network-for-continuous-sign-language | 18.3 |
| dense-temporal-convolution-network-for-sign | 36.5 |
| continuous-sign-language-recognition-through | 24.0 |
| self-mutual-distillation-learning-for | 20.5 |
| spatial-temporal-multi-cue-network-for | 20.7 |
| tcnet-continuous-sign-language-recognition | 18.9 |
| deep-radial-embedding-for-visual-sequence | 20.2 |