Jeu Educatif Multimodal d’Imitation Emotionnelle
Educational Multimodal Emotional Imitation Game
Interpersonal communication is based on a complex multimodal processing of emotional cues such as facial expressions and vocal intonations. Children with autism have great difficulty understanding and producing these socio-emotional cues.
JEMImE project aims to design new algorithms for multimodal emotion recognition to assess the quality of the emotions produced by children. The aim is to help children with autism learn to imitate and mimic facial and vocal emotions to express the appropriate expression in a given context.
From a technological standpoint, the algorithms must be able to analyze the spontaneous behavior of children in a natural environment. The challenge is to develop methods to both robust, real-time and multi-modal.
Programme ANR : Contenus numériques et interactions (CONTINT) 2013
Référence projet : ANR-13-CORD-0004
Partenaires : CoBTeK ISIR / LIRIS