Facial expression recognition using HessianMKL based multiclass-SVM

摘要: Multikernel learning (MKL) has recently received great attention in the field of computer vision and pattern recognition. The idea behind MKL is to optimally combine utilize multiple kernels features instead using a single kernel classifiers. This paper presents novel framework for problem by expanding HessianMKL algorithm into multiclass-SVM with one-against-one rule. Our learns one weight vector each binary classifier compared SimpleMKL based which jointly same all proposed method utilized recognize six basic facial expressions neutral expression combining three functions, RBF, Gaussian, polynomial function two image representations, HoG LBPH features. experimental results show that our performed better than SVM classifiers equipped type feature as well multiclass-SVM.