Contextualized End-to-end Automatic Speech Recognition with Intermediate Biasing Loss
Fuente:
arXiv
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866916388491231232 |
|---|---|
| author | Shakeel, Muhammad Sudo, Yui Peng, Yifan Watanabe, Shinji |
| author_facet | Shakeel, Muhammad Sudo, Yui Peng, Yifan Watanabe, Shinji |
| contents | Contextualized end-to-end automatic speech recognition has been an active research area, with recent efforts focusing on the implicit learning of contextual phrases based on the final loss objective. However, these approaches ignore the useful contextual knowledge encoded in the intermediate layers. We hypothesize that employing explicit biasing loss as an auxiliary task in the encoder intermediate layers may better align text tokens or audio frames with the desired objectives. Our proposed intermediate biasing loss brings more regularization and contextualization to the network. Our method outperforms a conventional contextual biasing baseline on the LibriSpeech corpus, achieving a relative improvement of 22.5% in biased word error rate (B-WER) and up to 44% compared to the non-contextual baseline with a biasing list size of 100. Moreover, employing RNN-transducer-driven joint decoding further reduces the unbiased word error rate (U-WER), resulting in a more robust network. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2406_16120 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Contextualized End-to-end Automatic Speech Recognition with Intermediate Biasing Loss Shakeel, Muhammad Sudo, Yui Peng, Yifan Watanabe, Shinji Audio and Speech Processing Computation and Language Sound Contextualized end-to-end automatic speech recognition has been an active research area, with recent efforts focusing on the implicit learning of contextual phrases based on the final loss objective. However, these approaches ignore the useful contextual knowledge encoded in the intermediate layers. We hypothesize that employing explicit biasing loss as an auxiliary task in the encoder intermediate layers may better align text tokens or audio frames with the desired objectives. Our proposed intermediate biasing loss brings more regularization and contextualization to the network. Our method outperforms a conventional contextual biasing baseline on the LibriSpeech corpus, achieving a relative improvement of 22.5% in biased word error rate (B-WER) and up to 44% compared to the non-contextual baseline with a biasing list size of 100. Moreover, employing RNN-transducer-driven joint decoding further reduces the unbiased word error rate (U-WER), resulting in a more robust network. |
| title | Contextualized End-to-end Automatic Speech Recognition with Intermediate Biasing Loss |
| topic | Audio and Speech Processing Computation and Language Sound |
| url | https://arxiv.org/abs/2406.16120 |