Face Recognition Based on Improved Multiscale Convolutional Neural Network
Shuang Liu, Xuemei Lei, Zheng Li · 2020
Image feature extraction using convolution neural network has been widely used in face recognition and other image recognition problems.A face recognition algorithm based on improved multi-scale convolutional neural network is proposed in this paper. The algorithm first performs histogram equalization on the input image to increase the contrast of the face image. The neural network structure only contains three convolution layers. The features extracted by convolution of each layer are fused and then input into Softmax for classification. The algorithm has few network layers and simple structure, and can fully extract face features. The recognition rates are 99.4% and 98.9% respectively in ORL and Yale's small face data sets.