Factors Related To The Improvement of Face Anti-Spoofing Detection Techniques With CNN Classifier

Abstract

Face recognition is one of the most successful application & has recently gain popularity with significant attention. Extensive research has been done in recognising the identity of the user from their facial image. Security issue on face recognition systems persists as a primary concern. although there are so many detection methods have been proposed but still it has some drawbacks in terms of parameters performance, size of datasets and generalisation ability to detect unseen face attacks So it is a challenging task to the researchers to proposed a robust face detection technique. This paper adopted comprehensive presentation of proposed Anti-spoofing techniques followed by features, datasets, parameters. Paper also provides experimental view on extensive comparative analysis of parameters, classifiers and databases which will be use to protect from various types of Face Spoofing attacks and depicted the purely CNN based existing methodology with general Face Spoofing detection module.