access_time July 10, 2020 at 10:00AM até July 10, 2020 at 11:30AM
place https://videoconf-colibri.zoom.us/j/92491584720 Password: 101648 (SSO authentication required)
Music is omnipresent in human life due to its many functions. In order to better harness its power, its semantics need to be studied, so that computational models can be equipped with a more human-centered understanding. Music conveys meaning via metaphorical and transmodal mappings, as supported by embodied cognition research. These inferences are realized in the brain as neural circuits and can even manifest overtly as motor activity, such as in dance. I claim that, in addition to music semantics being represented this way, generic semantic vector spaces can be learned, by correlating listener perceptual responses and interpretations. In this presentation, I review theoretical aspects regarding semantics and cognition, and I also present experimental results validating and leveraging some theoretical intuitions via predictive computational models of music semantics.
local_offer CAT exam
person Candidato: Francisco Raposo
supervisor_account Orientador: Prof. David Matos