Saved in:
Bibliographic Details
Main Authors: Yang, Hongli, Peng, Yizhou, Huang, Hao, Li, Sheng
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2506.21576
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912607333515264
author Yang, Hongli
Peng, Yizhou
Huang, Hao
Li, Sheng
author_facet Yang, Hongli
Peng, Yizhou
Huang, Hao
Li, Sheng
contents Large-scale multilingual ASR models like Whisper excel in high-resource settings but face challenges in low-resource scenarios, such as rare languages and code-switching (CS), due to computational costs and catastrophic forgetting. We explore Soft Prompt Tuning (SPT), a parameter-efficient method to enhance CS ASR while preserving prior knowledge. We evaluate two strategies: (1) full fine-tuning (FFT) of both soft prompts and the entire Whisper model, demonstrating improved cross-lingual capabilities compared to traditional methods, and (2) adhering to SPT's original design by freezing model parameters and only training soft prompts. Additionally, we introduce SPT4ASR, a combination of different SPT variants. Experiments on the SEAME and ASRU2019 datasets show that deep prompt tuning is the most effective SPT approach, and our SPT4ASR methods achieve further error reductions in CS ASR, maintaining parameter efficiency similar to LoRA, without degrading performance on existing languages.
format Preprint
id arxiv_https___arxiv_org_abs_2506_21576
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Adapting Whisper for Parameter-efficient Code-Switching Speech Recognition via Soft Prompt Tuning
Yang, Hongli
Peng, Yizhou
Huang, Hao
Li, Sheng
Computation and Language
Artificial Intelligence
Sound
Audio and Speech Processing
Large-scale multilingual ASR models like Whisper excel in high-resource settings but face challenges in low-resource scenarios, such as rare languages and code-switching (CS), due to computational costs and catastrophic forgetting. We explore Soft Prompt Tuning (SPT), a parameter-efficient method to enhance CS ASR while preserving prior knowledge. We evaluate two strategies: (1) full fine-tuning (FFT) of both soft prompts and the entire Whisper model, demonstrating improved cross-lingual capabilities compared to traditional methods, and (2) adhering to SPT's original design by freezing model parameters and only training soft prompts. Additionally, we introduce SPT4ASR, a combination of different SPT variants. Experiments on the SEAME and ASRU2019 datasets show that deep prompt tuning is the most effective SPT approach, and our SPT4ASR methods achieve further error reductions in CS ASR, maintaining parameter efficiency similar to LoRA, without degrading performance on existing languages.
title Adapting Whisper for Parameter-efficient Code-Switching Speech Recognition via Soft Prompt Tuning
topic Computation and Language
Artificial Intelligence
Sound
Audio and Speech Processing
url https://arxiv.org/abs/2506.21576