Facial Expression Recognition using Feature Extraction with Hybrid KFDA and CLBP
Facial Expression gives vital information about the emotion of human beings. It is a speedily growing and evergreen research field in the sector of computer vision, artificial intelligence and automation. We are proposing a facial expression recognition model that analyzes an image of a face and detects alignments to six different emotions which are Anger, Disgust, Fear, Happy and Neutral. This work focuses on the study of computer vision strategies intended to improve the accuracy of recognition and computational effectiveness through the application of certain alterations in terms of face location, feature extraction, and classifications algorithms. In this framework, the Kernel Fisher discriminant analysis (KFDA) and the Compound Local binary pattern (CLBP) are used for classifying features using the SVM. LS-SVM expresses the training in terms of solving a set of linear equations instead of quadratic programming as for the standard SVM case. An iterative training algorithm for LS-SVM based on a conjugate gradient method is then applied.
Keywords: Face recognition, KPCA analysis, compound Local binary pattern, SVM, KFDA.