Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2401.03689 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866912180546306048 |
|---|---|
| author | Liu, Wei Hou, Jingyong Yang, Dong Cao, Muyong Lee, Tan |
| author_facet | Liu, Wei Hou, Jingyong Yang, Dong Cao, Muyong Lee, Tan |
| contents | Toward high-performance multilingual automatic speech recognition (ASR), various types of linguistic information and model design have demonstrated their effectiveness independently. They include language identity (LID), phoneme information, language-specific processing modules, and cross-lingual self-supervised speech representation. It is expected that leveraging their benefits synergistically in a unified solution would further improve the overall system performance. This paper presents a novel design of a hierarchical information path, named LUPET, which sequentially encodes, from the shallow layers to deep layers, multiple aspects of linguistic and acoustic information at diverse granularity scales. The path starts from LID prediction, followed by acoustic unit discovery, phoneme sharing, and finally token recognition routed by a mixture-of-expert. ASR experiments are carried out on 10 languages in the Common Voice corpus. The results demonstrate the superior performance of LUPET as compared to the baseline systems. Most importantly, LUPET effectively mitigates the issue of performance compromise of high-resource languages with low-resource ones in the multilingual setting. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2401_03689 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | LUPET: Incorporating Hierarchical Information Path into Multilingual ASR Liu, Wei Hou, Jingyong Yang, Dong Cao, Muyong Lee, Tan Audio and Speech Processing Sound Toward high-performance multilingual automatic speech recognition (ASR), various types of linguistic information and model design have demonstrated their effectiveness independently. They include language identity (LID), phoneme information, language-specific processing modules, and cross-lingual self-supervised speech representation. It is expected that leveraging their benefits synergistically in a unified solution would further improve the overall system performance. This paper presents a novel design of a hierarchical information path, named LUPET, which sequentially encodes, from the shallow layers to deep layers, multiple aspects of linguistic and acoustic information at diverse granularity scales. The path starts from LID prediction, followed by acoustic unit discovery, phoneme sharing, and finally token recognition routed by a mixture-of-expert. ASR experiments are carried out on 10 languages in the Common Voice corpus. The results demonstrate the superior performance of LUPET as compared to the baseline systems. Most importantly, LUPET effectively mitigates the issue of performance compromise of high-resource languages with low-resource ones in the multilingual setting. |
| title | LUPET: Incorporating Hierarchical Information Path into Multilingual ASR |
| topic | Audio and Speech Processing Sound |
| url | https://arxiv.org/abs/2401.03689 |