FAREC — CNN based efficient face recognition technique using Dlib

摘要

Despite of advancement in face recognition, it has received much more attention in last few decades in the field of research and in commercial markets this project proposes an efficient technique for face recognition system based on Deep Learning using Convolutional Neural Network (CNN) with Dlib face alignment. The paper describes the process involved in the face recognition like face alignment and feature extraction. The paper also emphasizes the importance of the face alignment, thus the accuracy and False Acceptance Rate (FAR) is observed by using proposed technique. The computational analysis shows the better performance than other state-of-art approaches. The work has been done on Face Recognition Grand challenge (FRGC) dataset and giving accuracy of 96% with FAR of 0.1.