Keyword-Guided Adaptation of Automatic Speech Recognition

TL;DR

This paper introduces a keyword-guided adaptation method for Whisper-based ASR models, improving jargon recognition and overall accuracy, especially in noisy environments and unseen languages.

Contribution

The paper presents two novel approaches, KG-Whisper and KG-Whisper-PT, for dynamically guiding ASR models using keyword prompts to enhance recognition accuracy.

Findings

Significant improvement in keyword recognition accuracy.

Average WER reduction of 5.1% in unseen languages.

Effective steering of the decoder towards specified prompts.

Abstract

Automatic Speech Recognition (ASR) technology has made significant progress in recent years, providing accurate transcription across various domains. However, some challenges remain, especially in noisy environments and specialized jargon. In this paper, we propose a novel approach for improved jargon word recognition by contextual biasing Whisper-based models. We employ a keyword spotting model that leverages the Whisper encoder representation to dynamically generate prompts for guiding the decoder during the transcription process. We introduce two approaches to effectively steer the decoder towards these prompts: KG-Whisper, which is aimed at fine-tuning the Whisper decoder, and KG-Whisper-PT, which learns a prompt prefix. Our results show a significant improvement in the recognition accuracy of specified keywords and in reducing the overall word error rates. Specifically, in unseen language generalization, we demonstrate an average WER improvement of 5.1% over Whisper.