A unified theoretical Bayesian model of speech communication

Moulin-Frier, Clément and Schwartz, Jean-Luc and Diard, Julien and Bessière, Pierre

Link to the article


Based on a review of models and theories in speech communication, this paper proposes an original Bayesian framework able to express each of them in a unified way. This framework allows to selectively incorporate motor processes in perception or auditory representations in production, thus implementing components of a perceptuo-motor link in speech communication processes. This provides a basis for future computational works on the joint study of perception, production and their coupling in speech communication.


  title = {A unified theoretical Bayesian model of speech communication},
  author = {{Moulin-Frier}, Cl{\'e}ment and Schwartz, {Jean-Luc} and Diard, Julien and Bessi{\`e}re, Pierre},
  booktitle = {1st conference on Applied Digital Human Modeling, Miami, USA},
  year = {2010},
  keywords = {Speech Communication, Cognitive Bayesian Modeling, Sensory-Motor interaction},
  hal_id = {hal-01059208}