Investigation of Data Augmentation Techniques for Disordered Speech Recognition
This work addresses the problem of limited data for developing speech recognition systems for people with speech disorders, representing an incremental improvement.
The paper tackled the challenge of disordered speech recognition by investigating data augmentation techniques, resulting in a 2.92% absolute (9.3% relative) reduction in word error rate and an overall WER of 26.37% on a test set of 16 dysarthric speakers.
Disordered speech recognition is a highly challenging task. The underlying neuro-motor conditions of people with speech disorders, often compounded with co-occurring physical disabilities, lead to the difficulty in collecting large quantities of speech required for system development. This paper investigates a set of data augmentation techniques for disordered speech recognition, including vocal tract length perturbation (VTLP), tempo perturbation and speed perturbation. Both normal and disordered speech were exploited in the augmentation process. Variability among impaired speakers in both the original and augmented data was modeled using learning hidden unit contributions (LHUC) based speaker adaptive training. The final speaker adapted system constructed using the UASpeech corpus and the best augmentation approach based on speed perturbation produced up to 2.92% absolute (9.3% relative) word error rate (WER) reduction over the baseline system without data augmentation, and gave an overall WER of 26.37% on the test set containing 16 dysarthric speakers.