On the Use of Multi-attribute Decision Making for Combining Audio-Lingual and Visual-Facial Modalities in Emotion Recognition

摘要: In this chapter, we present and discuss a novel approach that have developed for the integration of audio-lingual visual-facial modalities in bi-modal user interface affect recognition. Even though researchers acknowledge two can provide information is complementary to each other with respect recognition, satisfactory progress has not yet been achieved towards modalities. our research reported herein, combination from perspective human observer by employing multi-criteria decision making theory dynamic recognition computer users. Our includes specification strengths weaknesses modality concerning 6 basic emotion states, namely happiness, sadness, surprise, anger disgust, as well emotionless state which refer neutral. We empirical studies conducted involving users observers emotions Based on results studies, assign weights criteria application theory. Additionally, may be used field currently unavailable relevant literature.