Uniformed two Local Binary Pattern Combined with Neighboring Support Vector Classifier for Classification
Mohammed Ngadi, Aouatif Amine, Bouchra Nassih, Hanaâ Hachimi, A. El-Attar · 2017
Due to its numerous real life applications, face classification has been one of the main research topics in computer vision and machine learning in recent years. In this paper, we develop an efficient and practical method for face classification. Our approch (LBP u2-NSVC) belongs to hybrid methods; it’s based on the combination of Uniform Local Binary Patterns (LBPu2) and a the classifier Neighboring Support Vector Classifier (NSVC). To be very simple, given a face dataset, we start by computing the LBP features first. Later on, these features are used to train a classifier using the new NSVC algorithm. Our experiments show that by using this approach and tuning well the NSVC parameters, we can get better performances than state-of-the-art similar algorithms. To confirm this, we use the famous MIT-CBCL face dataset for the different tests.