Artificial Neural Network Based Ensemble Approach for Multicultural Facial Expressions Analysis

Facial expressions convey exhaustive information about human emotions and the most interactive way of social collaborations, despite differences in ethnicity, culture, and geography. Due to cultural differences, the variations in facial structure, facial appearance, and facial expression representation are the main challenges to the facial expression recognition system. These variations necessitate the need for multicultural facial expression analysis. This study presents several computational algorithms to handle these variations to get high expression recognition accuracy. We propose an artificial neural network-based ensemble classifier for multicultural facial expression analysis. The facial images from the Japanese female facial expression database, Taiwanese facial expression image database, and RadBoud faces database are combined to form a multi-culture facial expression dataset. The participants in the multicultural dataset originate from four ethnic regions including Japanese, Taiwanese, Caucasians, and Moroccans. Local binary pattern, uniform local binary pattern, and principal component analysis are applied for facial feature representation. Experimental results prove that facial expressions are innate and universal across all cultures with minor variations.