Exploring Generative Error Correction for Dysarthric Speech Recognition
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866908380278292480 |
|---|---|
| author | La Quatra, Moreno Koudounas, Alkis Salerno, Valerio Mario Siniscalchi, Sabato Marco |
| author_facet | La Quatra, Moreno Koudounas, Alkis Salerno, Valerio Mario Siniscalchi, Sabato Marco |
| contents | Despite the remarkable progress in end-to-end Automatic Speech Recognition (ASR) engines, accurately transcribing dysarthric speech remains a major challenge. In this work, we proposed a two-stage framework for the Speech Accessibility Project Challenge at INTERSPEECH 2025, which combines cutting-edge speech recognition models with LLM-based generative error correction (GER). We assess different configurations of model scales and training strategies, incorporating specific hypothesis selection to improve transcription accuracy. Experiments on the Speech Accessibility Project dataset demonstrate the strength of our approach on structured and spontaneous speech, while highlighting challenges in single-word recognition. Through comprehensive analysis, we provide insights into the complementary roles of acoustic and linguistic modeling in dysarthric speech recognition |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2505_20163 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Exploring Generative Error Correction for Dysarthric Speech Recognition La Quatra, Moreno Koudounas, Alkis Salerno, Valerio Mario Siniscalchi, Sabato Marco Computation and Language Audio and Speech Processing Despite the remarkable progress in end-to-end Automatic Speech Recognition (ASR) engines, accurately transcribing dysarthric speech remains a major challenge. In this work, we proposed a two-stage framework for the Speech Accessibility Project Challenge at INTERSPEECH 2025, which combines cutting-edge speech recognition models with LLM-based generative error correction (GER). We assess different configurations of model scales and training strategies, incorporating specific hypothesis selection to improve transcription accuracy. Experiments on the Speech Accessibility Project dataset demonstrate the strength of our approach on structured and spontaneous speech, while highlighting challenges in single-word recognition. Through comprehensive analysis, we provide insights into the complementary roles of acoustic and linguistic modeling in dysarthric speech recognition |
| title | Exploring Generative Error Correction for Dysarthric Speech Recognition |
| topic | Computation and Language Audio and Speech Processing |
| url | https://arxiv.org/abs/2505.20163 |