Ramón y Cajal fellow at TALN Group
Head of the Expressive Speech Lab
Coordinator of the Master in Intelligent Interactive Systems
Why is synthetic speech still so boring? What can we learn from prosody?
Speech synthesis still sufferst from a lack of naturalness, because human speech naturalness and expressiveness is conveyed through many linguistic dimensions that are not considered in text-to-speech systems. I am passionate about analysing how linguistic structures relate to prosody, in order to understand the role that prosodic elements take in spoken discourse, and to transfer such information into several fields: speech synthesis, automatic speech recognition, and the use of prosody to detect mental and physical disorders.
Department of Information and Communication Technologies
Tànger building (Poblenou campus)
Tànger, 122-140
08018 Barcelona
(+34) 93 542 2569