Measuring the intelligibility of dysarthric speech through automatic speech recognition in a pluricentric language

Speech intelligibility is an essential though complex construct for evaluating dysarthric speech. Various procedures can be used to measure speech intelligibility, most of which are based on subjective ratings assigned by experts. Since these and laborious, automatic recognition (ASR) has been proposed obtain objective metrics intelligibility. Although promising results have reported, ASR generally requires large amounts data consisting recorded annotated In the present study, we explored possibility using resources from dominant language variety improve performance systems non-dominant same pluricentric language. Dutch as example a language, with Netherlandic considered Flemish variety. The evaluated two types metrics: orthographic transcriptions global assessments, both obtained Overall, show that contribute improving developing objective, measures only when no available training models.