Keynote Speaker

Talk Title: Voice puppetry: exploring dramatic performance to develop speech synthesis

Talk Abstract:State-of-the-art speech synthesis owes much to modern AI machine learning, with recurrent neural networks becoming the new standard. However, how you say something is just as important as what you say. If we draw inspiration from human dramatic performance, ideas such as artistic direction can help us design interactive speech synthesis systems which can be finely controlled by a human voice. This “voice puppetry” has many possible applications from film dubbing to the pre-creation of prompts for a conversational agent. Previous work in voice puppetry has raised the question of how such a system should work and how we might interact with it. In this talk I describe a prototype voice puppetry system, share audio generated from it and discuss the issues in using such a system. I will give an example of a script follower system that can be used as a test harness for performance like content and talk about its impact at the Glasgow Science Festival.

Brief Bio:  Matthew Aylett is an Associate Professor in Applied AI at Heriot Watt and co-founder and Chief Scientific Officer of CereProc, a speech synthesis company based in Edinburgh. He holds the titles of Royal Society Industrial Fellowship alumnus and Honorary Fellow at the School of Informatics, at the University of Edinburgh. He received his PhD and MSc in Speech and Language Technology from the University of Edinburgh and holds a BAHons in Computing and Artificial Intelligence from the University of Sussex.

Dr Aylett also serves as the Deputy Chief Scientist of the Scott Morgan Foundation(SMF) where they develop assistive technologies to empower people with disabilities such as Motor Neuron Disease (MND) and other life-challenging conditions to lead full and socially rich lives. The goal of SMF is to use human-centered AI and cutting-edge technology to improve accessibility for disabled people, thus transforming their lives for the better.