Corpus Design and Automatic Speech Recognition for Deaf and Hard-of-Hearing People
Akio Kobayashi, Keiichi Yasu, Hiromitsu Nishizaki, Norihide Kitaoka · 2021 IEEE 10th Global Conference on Consumer Electronics (GCCE) · 2021
This study describes automatic speech recognition (ASR) for the deaf and hard-of-hearing people. In the relevant literature, ASR for the deaf has been studied in a manner similar to the recognition of speech by people with dysarthria. However, prior studies have been conducted over a small number of deaf speakers. Therefore, to date, it remains unclear how the performance of ASR varies with different speakers. We conducted phoneme recognition experiments using speech from 12 deaf students to obtain analytical results from the perspective of ASR performance.