Evaluating performance of commercial automatic speech recognition systems for speakers with dysarthria
Year: 2022
Type: article
Abstract: Dysarthria refers to motor neuron speech disorders that limit the ability to control muscle groups used for speech production. Dysarthric speech has reduced intelligibility due to being slurry, breathy and slow compared to typical speech. The modern advancements of automatic speech recognition (ASR) and the increase of speechcontrolled systems raise the question on their inclusiveness for speakers with dysarthria. METHOD RESULTS DISCUSSIONAcknowledgement This study was funded by AIT Presidents Doctoral Scholarship This study evaluated the performance of two popular commercial ASR systems with dysarthric speech and compared the performance with state-of-the-art dysarthric speech ASR systems.
Source: Research@THEA
Authors: Fahim, Sirajuim, Murray, Niall, Flynn, Ronan
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7132302899
