Facial expression recognition based on Gabor feature and neural network
Lei Pang, Nianqiang Li, Li Zhao, Shi Wenxiu, Yunpan Du · 2018
With the development of human-computer interaction, emotional computing has gradually become a hot issue in computer vision research. Human expressions contain a wealth of information. How to make the computer fully extract facial expression information and understand human emotions is an urgent problem to be solved. The difficulty of facial expression recognition lies in facial expression extraction and facial expression classification. This paper analyzes the local features of facial expression extracted by Gabor wavelet transform and performs the various dimensional reduction on the problem of feature redundancy, which balances the feature dimension and The contribution rate of the feature. After extracting the features, a suitable BP neural network is constructed, and the extracted features are trained as a data set to the neural network to obtain a good facial expression classifier.