Comparing HMAX and BoVW Models for Large-Scale Image Classification

Image classification is one of the most important topics in computer vision. It became crucial for large image datasets. In literature, several approaches are proposed. this context, Bag-of-Visual Words (BoVW) model has been widely used. The BoVW relies on building visual vocabulary and images represented as histograms words. However, recently, attention shifted to use complex architectures which characterized by multilevel processing. HMAX (Hierarchical Max-pooling model) attracted a great deal classification, due its architecture, alternates layers feature extraction with pooling. This paper aims at comparing bags words using To achieve goal, we study features obtained SIFT (Scale-Invariant Feature Transform) descriptors, compare them features. performed support vector machine (SVM) classifiers. Both models tested ImageNet OpenImages datasets results have shown that performance outperforms model.