A study on feature extraction for face recognition using Self Organizing Maps

Abstract

This paper deals with the study related to the face recognition algorithms and process of developing a Self-Organizing Map (SOM) in order to carry out the process of face recognition in case of humans. Initially the paper deals with the various general steps involved in the structure and statistics based face recognition algorithms. However in the later part the key step used in the unsupervised algorithm as well as the combination of SOM and Hierarchical Self Organizing Map (HSOM) along with the aid of Gabor filters were discussed in order to carry out an efficient process of facial recognition. The feature selection criteria are also discussed in detail in order to achieve a high end result.

Keywords:SOM, HSOM, Gabor filters, unsupervised learning, feature extraction