Deep understanding and modelling of the hierarchical structure of Prosody

ended

The prime objective of the ProsoDeep project is to gain a deeper understanding of the language of prosody through the analysis of all the levels in the production hierarchy of prosody. In particular, it will exploit the benefits of the top-down and bottom-up approaches through their incorporation within a Deep Prosody Model (DPM). The DPM will facilitate the advancement of speech technologies that rely both on the synthesis of prosody, e.g. text-to-speech (TTS) systems, and its analysis, e.g. speech emotion recognition (SER). To reach this objective the project will draw on a variety of scientific fields, including signal processing, physiology of prosody production and biomechanics, linguistics, and machine learning, and will also be augmented with respiratory measurements.