Analýza robustnosti a efektivity metod klasifikace obličeje v rámci biometrické identifikace

Abstract

This thesis deals with the effectiveness and robustness of facial classification methods in the biometric identification. The subject of this thesis is the comparative analysis of neural networks for face detection and recognition from static images using the MATLAB simulation interface. In the first part of this thesis, there is a general overview of the methods used for face recognition. After that, the thesis is specifically focused to neural networks, which are the subject of investigation. This is followed by data acquisition for subsequent training analysis and evaluation of the output data. The last part of the thesis is dedicated to the proposal of possible improvements for the eventual continuation of the research.