A NOVEL BIOMETRIC APPROACH FOR FACIAL IMAGE RECOGNITION USING DEEP LEARNING TECHNIQUES

In this paper, age and gender specific user identity recognition is implemented using facial features for biometric application. A novel approach has been developed by extracting Local Binary Pattern (LBP) and Gray Level Co-Occurrence Matrix (GLCM) image features which effectively represents facial and skin regions of user to classify facial images based on various age groups and gender. These extracted features are classified by Convolution Neural Network (CNN), Region-based CNN (RCNN) and Fast RCNN, namely three popular Deep Learning Classification techniques using IMDB wikicrop facial dataset. Experimentations using CNN classifier achieved best result of 96.4% accuracy in contrast to other two classifier results.