Robust facial expression recognition based on RPCA and AdaBoost
Xia Mao, Yuli Xue, Zheng Li, Kang Huang, ShanWei Lv · 2009
In this paper, we consider the problem of robust facial expression recognition and propose a novel scheme for facial expression recognition under facial occlusion. There are two main contributions in this work. Firstly, a novel method for facial occlusion detection based on robust principal component analysis (RPCA) and saliency detection performs efficiently to detect facial occlusions. Secondly, a novel method based on occlusion reconstruction and reweighted AdaBoost classification is prosed for facial expression recognition. Experimental results have shown the effectiveness of our proposed method for robust facial expression recognition.