Synthetic Data Domain Adaptation for ASR via LLM-based Text and Phonetic Respelling Augmentation

TL;DR

This paper introduces a novel synthetic data augmentation framework for domain adaptation in end-to-end ASR, combining LLM-based text augmentation with phonetic respelling to improve robustness on domain-specific data.

Contribution

It presents a new phonetic respelling augmentation method and an LLM-based text augmentation pipeline for better domain adaptation in ASR.

Findings

Consistent WER reductions across four datasets.

Enhanced lexical diversity and pronunciation variability.

Improved robustness of ASR models on domain-specific data.

Abstract

End-to-end automatic speech recognition often degrades on domain-specific data due to scarce in-domain resources. We propose a synthetic-data-based domain adaptation framework with two contributions: (1) a large language model (LLM)-based text augmentation pipeline with a filtering strategy that balances lexical diversity, perplexity, and domain-term coverage, and (2) phonetic respelling augmentation (PRA), a novel method that introduces pronunciation variability through LLM-generated orthographic pseudo-spellings. Unlike conventional acoustic-level methods such as SpecAugment, PRA provides phonetic diversity before speech synthesis, enabling synthetic speech to better approximate real-world variability. Experimental results across four domain-specific datasets demonstrate consistent reductions in word error rate, confirming that combining domain-specific lexical coverage with realistic pronunciation variation significantly improves ASR robustness.