Speech Recognition on CoVoST 2 (test)
15.7WER (Fr)w2v2-CTC
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| w2v2-CTCLearning Paradigm=Supervised learning + pre-training, Decoding Method=4-gram language model2022.10 | 15.7 | 7 | 7.1 | 11.1 | 5.9 | 8.9 | |
| TransformerLearning Paradigm=Supervised learning2022.10 | 18.3 | 16 | 31.4 | 65.7 | 51.8 | 32.6 | |
| w2vu2-CTCLearning Paradigm=Unsupervised learning, Decoding Method=4-gram language model2022.10 | 23.2 | 10.3 | 15.7 | 17.6 | 14.8 | 15.7 |