Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Liu, Hexin, Zhang, Xiangyu, Zhang, Haoyang, Garcia, Leibny Paola, Khong, Andy W. H., Chng, Eng Siong, Watanabe, Shinji
Format: Preprint
Veröffentlicht: 2024
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2403.05887
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866915590204030976
author Liu, Hexin
Zhang, Xiangyu
Zhang, Haoyang
Garcia, Leibny Paola
Khong, Andy W. H.
Chng, Eng Siong
Watanabe, Shinji
author_facet Liu, Hexin
Zhang, Xiangyu
Zhang, Haoyang
Garcia, Leibny Paola
Khong, Andy W. H.
Chng, Eng Siong
Watanabe, Shinji
contents Code-switching (CS) refers to the switching of languages within a speech signal and results in language confusion for automatic speech recognition (ASR). To address language confusion, we propose a language alignment loss (LAL) that aligns acoustic features to pseudo-language labels learned from the ASR decoder during ASR training. This approach enables frame-level language identification without the need for frame-level language annotations. To further tackle the complex token alternatives for language modeling in bilingual scenarios, we propose to employ large language models via a generative error correction method. A linguistic hint, derived from LAL outputs and decoded hypotheses, is introduced to guide the prompting and enhance the LLM-based generative error correction for CS-ASR. The proposed methods are evaluated on the SEAME dataset and data from the ASRU 2019 Mandarin-English code-switching speech recognition challenge. The incorporation of the proposed language alignment loss improves CS-ASR performance for both hybrid CTC/attention and Whisper models on both datasets, with only a negligible increase in the number of parameters. This work also highlights the efficacy of language alignment loss in balancing primary-language-dominant bilingual data during training, with an 8.6% relative improvement on the ASRU dataset compared to the baseline model. Performance evaluation using large language models reveals the advantage of the linguistic hint by achieving 14.1% and 5.5% relative improvement on test sets of the ASRU and SEAME datasets, respectively.
format Preprint
id arxiv_https___arxiv_org_abs_2403_05887
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Aligning Speech to Languages to Enhance Code-switching Speech Recognition
Liu, Hexin
Zhang, Xiangyu
Zhang, Haoyang
Garcia, Leibny Paola
Khong, Andy W. H.
Chng, Eng Siong
Watanabe, Shinji
Audio and Speech Processing
Code-switching (CS) refers to the switching of languages within a speech signal and results in language confusion for automatic speech recognition (ASR). To address language confusion, we propose a language alignment loss (LAL) that aligns acoustic features to pseudo-language labels learned from the ASR decoder during ASR training. This approach enables frame-level language identification without the need for frame-level language annotations. To further tackle the complex token alternatives for language modeling in bilingual scenarios, we propose to employ large language models via a generative error correction method. A linguistic hint, derived from LAL outputs and decoded hypotheses, is introduced to guide the prompting and enhance the LLM-based generative error correction for CS-ASR. The proposed methods are evaluated on the SEAME dataset and data from the ASRU 2019 Mandarin-English code-switching speech recognition challenge. The incorporation of the proposed language alignment loss improves CS-ASR performance for both hybrid CTC/attention and Whisper models on both datasets, with only a negligible increase in the number of parameters. This work also highlights the efficacy of language alignment loss in balancing primary-language-dominant bilingual data during training, with an 8.6% relative improvement on the ASRU dataset compared to the baseline model. Performance evaluation using large language models reveals the advantage of the linguistic hint by achieving 14.1% and 5.5% relative improvement on test sets of the ASRU and SEAME datasets, respectively.
title Aligning Speech to Languages to Enhance Code-switching Speech Recognition
topic Audio and Speech Processing
url https://arxiv.org/abs/2403.05887