Towards Effective and Efficient Non-autoregressive Decoding Using Block-based Attention Mask
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2024
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866916375223599104 |
|---|---|
| author | Wang, Tianzi Xie, Xurong Li, Zhaoqing Hu, Shoukang Jin, Zengrui Deng, Jiajun Cui, Mingyu Hu, Shujie Geng, Mengzhe Li, Guinan Meng, Helen Liu, Xunying |
| author_facet | Wang, Tianzi Xie, Xurong Li, Zhaoqing Hu, Shoukang Jin, Zengrui Deng, Jiajun Cui, Mingyu Hu, Shujie Geng, Mengzhe Li, Guinan Meng, Helen Liu, Xunying |
| contents | This paper proposes a novel non-autoregressive (NAR) block-based Attention Mask Decoder (AMD) that flexibly balances performance-efficiency trade-offs for Conformer ASR systems. AMD performs parallel NAR inference within contiguous blocks of output labels that are concealed using attention masks, while conducting left-to-right AR prediction and history context amalgamation between blocks. A beam search algorithm is designed to leverage a dynamic fusion of CTC, AR Decoder, and AMD probabilities. Experiments on the LibriSpeech-100hr corpus suggest the tripartite Decoder incorporating the AMD module produces a maximum decoding speed-up ratio of 1.73x over the baseline CTC+AR decoding, while incurring no statistically significant word error rate (WER) increase on the test sets. When operating with the same decoding real time factors, statistically significant WER reductions of up to 0.7% and 0.3% absolute (5.3% and 6.1% relative) were obtained over the CTC+AR baseline. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2406_10034 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Towards Effective and Efficient Non-autoregressive Decoding Using Block-based Attention Mask Wang, Tianzi Xie, Xurong Li, Zhaoqing Hu, Shoukang Jin, Zengrui Deng, Jiajun Cui, Mingyu Hu, Shujie Geng, Mengzhe Li, Guinan Meng, Helen Liu, Xunying Sound Artificial Intelligence Audio and Speech Processing This paper proposes a novel non-autoregressive (NAR) block-based Attention Mask Decoder (AMD) that flexibly balances performance-efficiency trade-offs for Conformer ASR systems. AMD performs parallel NAR inference within contiguous blocks of output labels that are concealed using attention masks, while conducting left-to-right AR prediction and history context amalgamation between blocks. A beam search algorithm is designed to leverage a dynamic fusion of CTC, AR Decoder, and AMD probabilities. Experiments on the LibriSpeech-100hr corpus suggest the tripartite Decoder incorporating the AMD module produces a maximum decoding speed-up ratio of 1.73x over the baseline CTC+AR decoding, while incurring no statistically significant word error rate (WER) increase on the test sets. When operating with the same decoding real time factors, statistically significant WER reductions of up to 0.7% and 0.3% absolute (5.3% and 6.1% relative) were obtained over the CTC+AR baseline. |
| title | Towards Effective and Efficient Non-autoregressive Decoding Using Block-based Attention Mask |
| topic | Sound Artificial Intelligence Audio and Speech Processing |
| url | https://arxiv.org/abs/2406.10034 |