Facial expression recognition using encoded dynamic features,” CVPR

In this paper, we propose a novel framework for video-based facial expression recognition, which can handle the data with various time resolution including a single frame. We first use the haar-like features to represent facial ap-pearance, due to their simplicity and effectiveness. Then we perform K-Means clustering on the facial appearance features to explore the intrinsic temporal patterns of each expression. Based on the temporal pattern models, we fur-ther map the facial appearance variations into dynamic bi-nary patterns. Finally, boosting learning is performed to construct the expression classifiers. Compared to previous work, the dynamic binary patterns encode the intrinsic dy-namics of expression, and our method makes no assumption on the time resolution of the data. Extensive experiments carried on the Cohn-Kanade database show the promising performance of the proposed method. 1