Journal of Soft Computing and Artificial Intelligence
Yazarlar: Mahmut DİRİK, Oscar CASTILLO, A. Fatih KOCAMAZ
Konular:Bilgisayar Bilimleri, Yapay Zeka
Anahtar Kelimeler:Action Unit,Emotion Recognition,Facial Expression Recognition,Human–Computer Interaction,Interval Type-2 Fuzzy System
Özet: Automatic recognition of facial emotion plays an effective and important role in Human–Computer Interaction (HCI). There are various emotion recognition approaches have been proposed in the literature. The analytic face model consisted of a 26-dimensional geometric feature vector. These properties are used effectively to identify facial changes resulting from different expressions. The variation and uncertainties of these features make the emotion recognition problem more complicated. For decreasing these complications, we propose a distance-based clustering and uncertainty measures of the base new method for Emotion Recognition from Facial Expression using automatically selects 19 diagnostics of Action Units (AUs) in a 2D facial image using Type-2 Fuzzy inference system. The proposed system includes an automated generation scheme of the geometric facial feature vector. The proposed system has classified six facial expressions using the MUG Facial Expression database. The experimental results show that the proposed model is very efficient in uncertainty management policy and recognizes six basic emotions with an average precision rate of 86.175%.