Facial Expression Recognition Based On Feature Extraction Of Key Sub-regions

In recent years,artificial intelligence technology has made great strides.As a research direction in the field of artificial intelligence,facial expression recognition has a huge application prospect in the field of human-computer interaction.At present,the research on expression recognition technology is more and more in-depth,but the accurate and rapid facial expression recognition is still challenging.By reading a large number of relevant literature,the existing problems in expression recognition research are analyzed and discussed.In order to improve the detection efficiency and precision of facial key points,select and make full use of key sub-regions and enhance facial expression feature description ability,this paper designs a facial expression recognition system based on the key sub-region feature extraction,and studies its key technologies.The main work of this paper is as follows:Aiming at the problem of detecting the key points in the face,this paper first uses the simplified deformable component model(DPM)to locate the key points on the face,so as to reduce the time complexity of DPM;Then the initial location information of the key points is assigned to the AAM model and fitted;Ultimately more quickly and accurately detected 68 facial key points.Aiming at the selection of key sub-regions,in this paper,based on the detection of facial key points,the key sub-regions are selected according to the facial motion coding system(FACS)and the saliency principle.Then the non-critical sub-region pixels are zeroed,which not only retains the overall structure of the face but also removes the influence of redundant information.Aiming at the problem of feature extraction,an improved local gradient coding operator(LMGC-HD)is proposed in this paper.The improved operator has a lower dimension,more fully describes the local deformation and is less affected by random noise and edge changes.Finally,this paper constructs a facial expression recognition experimental system based on the key sub-region feature extraction.In this paper,we use CK + data set to conduct experiments and use support vector machine(SVM)to classify and recognize.Experimental results show that our system can effectively improve the facial expression recognition rate.