Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2409.16469 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866916410669662208 |
|---|---|
| author | Velikovich, Leonid Li, Christopher Caseiro, Diamantino Kumar, Shankar Rondon, Pat Joshi, Kandarp Velez, Xavier |
| author_facet | Velikovich, Leonid Li, Christopher Caseiro, Diamantino Kumar, Shankar Rondon, Pat Joshi, Kandarp Velez, Xavier |
| contents | For end-to-end Automatic Speech Recognition (ASR) models, recognizing personal or rare phrases can be hard. A promising way to improve accuracy is through spelling correction (or rewriting) of the ASR lattice, where potentially misrecognized phrases are replaced with acoustically similar and contextually relevant alternatives. However, rewriting is challenging for ASR models trained with connectionist temporal classification (CTC) due to noisy hypotheses produced by a non-autoregressive, context-independent beam search.
We present a finite-state transducer (FST) technique for rewriting wordpiece lattices generated by Transformer-based CTC models. Our algorithm performs grapheme-to-phoneme (G2P) conversion directly from wordpieces into phonemes, avoiding explicit word representations and exploiting the richness of the CTC lattice. Our approach requires no retraining or modification of the ASR model. We achieved up to a 15.2% relative reduction in sentence error rate (SER) on a test set with contextually relevant entities. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2409_16469 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Spelling Correction through Rewriting of Non-Autoregressive ASR Lattices Velikovich, Leonid Li, Christopher Caseiro, Diamantino Kumar, Shankar Rondon, Pat Joshi, Kandarp Velez, Xavier Computation and Language Sound Audio and Speech Processing For end-to-end Automatic Speech Recognition (ASR) models, recognizing personal or rare phrases can be hard. A promising way to improve accuracy is through spelling correction (or rewriting) of the ASR lattice, where potentially misrecognized phrases are replaced with acoustically similar and contextually relevant alternatives. However, rewriting is challenging for ASR models trained with connectionist temporal classification (CTC) due to noisy hypotheses produced by a non-autoregressive, context-independent beam search. We present a finite-state transducer (FST) technique for rewriting wordpiece lattices generated by Transformer-based CTC models. Our algorithm performs grapheme-to-phoneme (G2P) conversion directly from wordpieces into phonemes, avoiding explicit word representations and exploiting the richness of the CTC lattice. Our approach requires no retraining or modification of the ASR model. We achieved up to a 15.2% relative reduction in sentence error rate (SER) on a test set with contextually relevant entities. |
| title | Spelling Correction through Rewriting of Non-Autoregressive ASR Lattices |
| topic | Computation and Language Sound Audio and Speech Processing |
| url | https://arxiv.org/abs/2409.16469 |