An Analysis of Personalized Speech Recognition System Development for the Deaf and Hard-of-Hearing

TL;DR

This paper evaluates the development of personalized speech recognition systems for deaf and hard-of-hearing individuals, demonstrating that collecting around 1000 utterances significantly improves model performance, with data augmentation techniques aiding smaller datasets.

Contribution

It provides a thorough analysis of personalized ASR for DHH speakers, including dataset collection, training frameworks, and practical recommendations for effective system development.

Findings

1000 utterances improve ASR performance significantly

Data augmentation helps when fewer than 1000 utterances are available

Personalized ASR systems can be efficiently developed with minimal data

Abstract

Deaf or hard-of-hearing (DHH) speakers typically have atypical speech caused by deafness. With the growing support of speech-based devices and software applications, more work needs to be done to make these devices inclusive to everyone. To do so, we analyze the use of openly-available automatic speech recognition (ASR) tools with a DHH Japanese speaker dataset. As these out-of-the-box ASR models typically do not perform well on DHH speech, we provide a thorough analysis of creating personalized ASR systems. We collected a large DHH speaker dataset of four speakers totaling around 28.05 hours and thoroughly analyzed the performance of different training frameworks by varying the training data sizes. Our findings show that 1000 utterances (or 1-2 hours) from a target speaker can already significantly improve the model performance with minimal amount of work needed, thus we recommend researchers to collect at least 1000 utterances to make an efficient personalized ASR system. In cases where 1000 utterances is difficult to collect, we also discover significant improvements in using previously proposed data augmentation techniques such as intermediate fine-tuning when only 200 utterances are available.