Community oriented wearable computing systems (COMMON-WEARS)

Overview:

In this project, we will develop a speech recogniser to improve well-being and communication in subjects who are frail due to laryngeal diseases. We will employ innovative artificial intelligence techniques to recognise the speech of a sample of subjects for whom current technologies are ineffective. This phase involves the collection of suitable voice samples, their characterisation, and the development of a basic machine learning model that recognises voices lacking regular harmonic structure. Recognition of pathological voices will give fragile individuals more inclusive access to a whole range of systems that have achieved ‘pervasive’ deployment in the most diverse application domains where voice is recognised and transformed for control, actuation and communication.