Welcome to Speech Prosody 2020!!

SP2020 will connect two aspects of prosody: the social and the scientific/technical. In our globalizing world, we have increased opportunities to talk with people from other societies, cultures, languages, religions, etc., where we face a large diversity of communication and interaction styles. Speech prosody plays a uniquely essential role for mutual understanding, impelling us to broaden and deepen our knowledge. On the scientific/technical side, the last decade has seen greatly improved performance in many prosody-related tasks, largely obtained by applying deep neural network models to enormous data sets, yet, ironically, as these network architectures become deeper, the linguistic knowledge they exploit becomes shallower. Speech Prosody 2020 will be a venue for fundamental work in human perception and production of prosody, for studies of prosody in different styles, and for knowledge-rich investigations and models of any aspect of prosody.  While technology pursues performance but science aims at interpretation, ultimately the two endeavors must proceed in concert.  Speech Prosody 2020 aims to deepen interactive communication among scientists and engineers, and specifically, to explore communicative and interactive prosody from all perspectives.

See you all in Tokyo
Nobuaki MINEMATSU