Abstract
Multi-culture facial expression recognition remains challenging due to cross cultural variations in facial expressions representation, caused by facial structure variations and culture specific facial characteristics. In this research, a joint deep learning approach called racial identity aware deep convolution neural network is developed to recognize the multicultural facial expressions. In the proposed model, a pre-trained racial identity network learns the racial features. Then, the racial identity aware network and racial identity network jointly learn the racial identity aware facial expressions. By enforcing the marginal independence of facial expression and racial identity, the proposed joint learning approach is expected to be purer for the expression and be robust to facial structure and culture specific facial characteristics variations. For the reliability of the proposed joint learning technique, extensive experiments were performed with racial identity features and without racial identity features. Moreover, culture wise facial expression recognition was performed to analyze the effect of inter-culture variations in facial expression representation. A large scale multi-culture dataset is developed by combining the four facial expression datasets including JAFFE, TFEID, CK+ and RaFD. It contains facial expression images of Japanese, Taiwanese, American, Caucasian and Moroccan cultures. We achieved 96% accuracy with racial identity features and 93% accuracy without racial identity features.
Highlights
Introduction published maps and institutional affilThe aim of a facial expression recognition system is to recognize the discrete categories of facial expressions such as happy, sad, surprise, angry, neutral, fear, contempt and disgust from still images or video sequences
A racial identity network is trained to learn the cultural variations in the multicultural dataset
Compiled with Racial Identity Aware (RIA-Deep Convolutional Neural Network (DCNN)) Facial Expressions Recognition
Summary
Introduction published maps and institutional affilThe aim of a facial expression recognition system is to recognize the discrete categories of facial expressions such as happy, sad, surprise, angry, neutral, fear, contempt and disgust from still images or video sequences. Though several recent studies focus on image sequence based facial expression recognition tasks [1], still image based facial expression recognition remains a difficult problem due to the following three reasons. The facial structure variations among the subjects of different cultures make the classification task difficult in some cases [2]. Different subjects express their emotions in different ways due to facial appearance variation and facial biometric shapes [4]. It contains five representative faces, labeled as “Moroccan”, “Caucasian”,
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have