People: Jose A. Gonzalez (Sheffield), Phil D. Green (Sheffield), Roger K. Moore (Sheffield), Damian Murphy (York), Amelia Gully (York), Andy Bulpitt (Leeds), Duane Carey (Leeds), James Gilbert (Hull), Lam A. Cheah (Hull).
Partners: Sheffield University, Leeds University, York University and Hull University.
Every year, some 17500 people in Europe and North America lose the power of speech after undergoing a laryngectomy. Loss of speech is known to cause social isolation, feelings of loss of identity and can lead to clinical depression. Conventional methods (e.g. oesophageal speech, the electrolarynx and speech valves) used to restore speech have their limitations in terms of quality of speech, ease of use and, in the case of implanted speech valves, the requirement for frequent valve replacement.
In Silent Speech we are investigating a revolutionary way of restoring speech after laryngectomy. The idea is to predict, in real time, the vocal tract shape from measurements of the lips and tongue movements obtained using magnetic sensing. From the estimated vocal tract shapes, speech is synthesised by simulation of sound propagation in the vocal tract.
Ultimately, the method we are investigating will allow those patients to recover those original voices, as the speech synthesiser will be personalised to each patient by using MRI images of the patient’s vocal tract.
Silent Speech is funded by the White Rose Collaboration Fund programme.