A computational prosthesis for the speech impaired – Think Kent Video

As part of the University of Kent Think Kent series – a collection of YouTube videos celebrating research with international impact at the University – Professor Ian McLoughlin describes how advances in speech technology are being adapted to assist patients who experience speech impairment.

In this work in the School of Computing Medway, the latest deep learning techniques are being combined with speech processing research to deliver algorithms that will form the heart of a wearable computational prostheses to tackle speech impairment. This kind of assistive technology blends humans and computers together to augment only the precise aspects of impairment, while leaving other aspects of speech production unchanged. This means that the reconstructed voice is mainly natural and is not synthesised, something that is important in ensuring the personality and characteristics of the speaker are maintained. This world first from the group at Kent aims to return the power of speech to voice-impaired patients worldwide.

During an amazingly varied career, Professor Ian McLoughlin has designed digital systems for satellites (including the world’s first cluster computer in space), in the troposphere, in daily use by emergency services, perched on the lip of a volcano crater, embedded in mass-market consumer devices, used for rehabilitation for voice-loss patients, or for Chinese speech communications. He has 15 patents, four books and over 200 peer-reviewed papers. He has worked in industry and academia in China, Singapore, New Zealand and the United Kingdom, and is a Fellow of the IET.

The talk may be viewed below or on YouTube via the link: https://youtu.be/–Dp_0RikXw