IRCAM 2021 - ViaDialog - David Guennec

Title of the intervention :

Towards helpful, customer-specific Text-To-Speech synthesis

Abstract of the intervention :

The subject of automatic speech synthesis began to be democratised in the 1990s. Each of us has already had to deal with these automatic answering machine voices which made us all suffer at first. Today, however, progress in both language understanding and the acoustic quality of speech synthesis approaches has enabled us to make giant leaps forward and new speech services are now rapidly increasing in quality and capability with increasingly human and expressive voices.
In this presentation, we will briefly review recent progress in speech synthesis. After this introduction, we will discuss the topics related to the customisation of synthesised voices to the needs of the client; and this at several levels. Firstly, at the level of the main components of speech: language, speech style, language register and gender for example. Secondly, issues at the level of the utterance; prosodic for the most part (pitch manipulation, flow). Finally, we will discuss the subsidiary elements that need to be taken into consideration to best meet the needs of clients and end-users of synthetic voice in our ever-changing world.

Information about the speaker : 

Name: David GUENNEC
Mini bio: A researcher in computer science with a passion for the history of sound reproduction, David Guennec specialises in the field of new voice technologies. After a PhD on speech synthesis, he moved on to the creation of voice assistants integrating the entire voice reproduction chain; from speech recognition to synthesis and natural language understanding. Currently working at ViaDialog, he focuses mainly on speech synthesis and recognition.

Similar articles