Personalized Fine-Tuning with Controllable Synthetic Speech from LLM-Generated Transcripts for Dysarthric Speech Recognition

TL;DR

This paper introduces a method for improving dysarthric speech recognition by generating synthetic speech data with LLM-guided fine-tuning of TTS, combined with personalized adaptation techniques, leading to significant WER reductions.

Contribution

The study presents a novel approach integrating synthetic speech generation with personalized fine-tuning using parameter-efficient methods for dysarthric speech recognition.

Findings

x-vector personalization reduces WER

AdaLoRA adapters outperform full fine-tuning

Synthetic data improves recognition accuracy

Abstract

In this work, we present our submission to the Speech Accessibility Project challenge for dysarthric speech recognition. We integrate parameter-efficient fine-tuning with latent audio representations to improve an encoder-decoder ASR system. Synthetic training data is generated by fine-tuning Parler-TTS to mimic dysarthric speech, using LLM-generated prompts for corpus-consistent target transcripts. Personalization with x-vectors consistently reduces word error rates (WERs) over non-personalized fine-tuning. AdaLoRA adapters outperform full fine-tuning and standard low-rank adaptation, achieving relative WER reductions of ~23% and ~22%, respectively. Further improvements (~5% WER reduction) come from incorporating wav2vec 2.0-based audio representations. Training with synthetic dysarthric speech yields up to ~7% relative WER improvement over personalized fine-tuning alone.