Extending Whisper with prompt tuning to target-speaker ASR

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Ma, Hao, Peng, Zhiyuan, Shao, Mingjie, Li, Jing, Liu, Ju
Natura: Preprint
Pubblicazione: 2023
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866929206350315520
author Ma, Hao
Peng, Zhiyuan
Shao, Mingjie
Li, Jing
Liu, Ju
author_facet Ma, Hao
Peng, Zhiyuan
Shao, Mingjie
Li, Jing
Liu, Ju
contents Target-speaker automatic speech recognition (ASR) aims to transcribe the desired speech of a target speaker from multi-talker overlapped utterances. Most of the existing target-speaker ASR (TS-ASR) methods involve either training from scratch or fully fine-tuning a pre-trained model, leading to significant training costs and becoming inapplicable to large foundation models. This work leverages prompt tuning, a parameter-efficient fine-tuning approach, to extend Whisper, a large-scale single-talker ASR model, to TS-ASR. Variants of prompt tuning approaches along with their configurations are explored and optimized for TS-ASR.Experimental results show that prompt tuning can achieve performance comparable to state-of-the-art full training approaches while only requiring about 1\% of task-specific model parameters. Notably, the original Whisper's features, such as inverse text normalization and timestamp tagging, are retained in target-speaker ASR, keeping the generated transcriptions natural and informative.
format Preprint
id arxiv_https___arxiv_org_abs_2312_08079
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Extending Whisper with prompt tuning to target-speaker ASR
Ma, Hao
Peng, Zhiyuan
Shao, Mingjie
Li, Jing
Liu, Ju
Computation and Language
Sound
Audio and Speech Processing
Target-speaker automatic speech recognition (ASR) aims to transcribe the desired speech of a target speaker from multi-talker overlapped utterances. Most of the existing target-speaker ASR (TS-ASR) methods involve either training from scratch or fully fine-tuning a pre-trained model, leading to significant training costs and becoming inapplicable to large foundation models. This work leverages prompt tuning, a parameter-efficient fine-tuning approach, to extend Whisper, a large-scale single-talker ASR model, to TS-ASR. Variants of prompt tuning approaches along with their configurations are explored and optimized for TS-ASR.Experimental results show that prompt tuning can achieve performance comparable to state-of-the-art full training approaches while only requiring about 1\% of task-specific model parameters. Notably, the original Whisper's features, such as inverse text normalization and timestamp tagging, are retained in target-speaker ASR, keeping the generated transcriptions natural and informative.
title Extending Whisper with prompt tuning to target-speaker ASR
topic Computation and Language
Sound
Audio and Speech Processing
url https://arxiv.org/abs/2312.08079