Speech Recognition On Tuda
Métriques
Test WER
Résultats
Résultats de performance de divers modèles sur ce benchmark
Tableau comparatif
| Nom du modèle | Test WER |
|---|---|
| scribosermo-fast-speech-to-text-models-for | 10.2% |
| open-source-german-distant-speech-recognition | 20.5% |
| automatic-speech-recognition-in-german-a | 5.82% |
| Modèle 4 | 18.6% |
| open-source-automatic-speech-recognition-for | 14.4% |
| ims-speech-a-speech-to-text-tool | 12.0% |
| ctc-segmentation-of-large-corpora-for-german | 12.8% |
| open-source-german-distant-speech-recognition | 39.6% |
| improved-open-source-automatic-subtitling-for | 6.93% |