Affective computing - the ability of a system to recognize, understand and simulate human emotional intelligence - is one of the most dynamic fields of HCI - Human Computer Interaction. These characteristics find their applicability in those areas where it is necessary to extend traditional cognitive communication with emotional features. That is why, Computer Based Speech Therapy Systems (CBST), and especially those involving children with speech disorders, require this qualitative shift. So in this paper we propose an original emotional framework recognition as an extension for our previous developed system - Logomon. A fuzzy model is used in order to interpret the values of specific physiological parameters and to obtain the emotional state of the subject. Moreover, an experiment that indicates the emotion pattern (average fuzzy sets) for each therapeutic sequence is also presented. The obtained results encourage us to continue working on automatic emotion recognition and provide important clues regarding the future development of our CBST.
展开▼
