Synthetic Data Augmentation for Speech Recognition

Overview

This field studies how to create extra training examples for speech recognizers when labeled recordings are limited or fail to represent real-world speakers and conditions. Common approaches include text-to-speech synthesis, voice conversion and cloning, language-model-generated transcripts or phonetic variants, and acoustic transformations such as speed or pitch changes. Researchers apply these methods to settings including low-resource languages, accents, code-switching, conversation, and disordered speech, and evaluate them with word error rates and task-specific error measures. A key open tension is that more natural-sounding synthetic data does not always improve recognition: its value depends on how well it matches the target speech, what variation it adds, and how it is mixed with real recordings.

Latest papers

31 papers4 papers in the last four weeks, down 20% on the four weeks before. 0.0% of all new papers.

Mar 2, 2026

DARS: Dysarthria-Aware Rhythm-Style Synthesis for ASR Enhancement

Dysarthric speech exhibits abnormal prosody and significant speaker variability, presenting persistent challenges for automatic speech recognition (ASR). While text-to-speech (TTS)-based data augmentation has shown potential, existing methods often fail to accurately model the pathological rhythm and acoustic style of dysarthric speech. To address this, we propose DARS, a dysarthria-aware rhythm-style synthesis framework based on the Matcha-TTS architecture. DARS incorporates a multi-stage rhythm predictor optimized by contrastive preferences between normal and dysarthric speech, along with a dysarthric-style conditional flow matching mechanism, jointly enhancing temporal rhythm reconstruction and pathological acoustic style simulation. Experiments on the TORGO dataset demonstrate that DARS achieves a Mean Cepstral Distortion (MCD) of 4.29, closely approximating real dysarthric speech. Adapting a Whisper-based ASR system with synthetic dysarthric speech from DARS achieves a 54.22% relative reduction in word error rate (WER) compared to state-of-the-art methods, demonstrating the framework's effectiveness in enhancing recognition performance.