Face Recognition Combined with Gabor Wavelet and Lightweight Convolutional Neural Network

Abstract

The existing deep convolutional neural network model has achieved high accuracy in face recognition, but it has a large amount of calculation, high resource consumption, and loss of local features of the faces. Therefore, this paper proposes a method of combining Gabor wavelet and lightweight convolutional neural network, using the GMNet network which consists of convolutional layer (Gabor layer) that implements the function of Gabor filter on the basic elements of convolutional neural network and lightweight convolutional neural network to extract more discriminative face features. It can not only reduce the number of parameters and computational complexity, but also extract more distinguished face features to improve the accuracy of face recognition. The experimental results on the LFW, CFP-FP, and AgeDB-30 datasets show that the improved model can enhance the performance of face recognition and maintain good results when the illumination, posture, age and other change.