Linguistic-Paralinguistic Speech Synthesis

The goal of this project is for the candidate to research and prepare an SNSF Ambizione proposal on natural language processing (NLP) and text-to-speech (TTS) synthesis. For many years NLP and TTS research fields have been advancing in parallel, introducing novelties in the research area of speech and language processing. Nonetheless, modern TTS systems even though are capable of producing very high quality synthetic neutral speech or even emotional speech, they lack the ability of producing synthetic speech continuously adapted to the meaning and higher semantics conveyed through the text which is synthesized. The intention of the candidate is to bridge this gap between NLP and TTS by taking advantage of NLP advances and incorporate them in TTS technologies.

Hasler Stiftung (Hasler Foundation)
Dec 01, 2015
Feb 28, 2016