Title: Face recognition under low illumination based on convolutional neural network
Authors: Yani Zhu; Xuan Ni; Hui Wang; Ye Yao
Addresses: College of Computer Science and Technology, Zhejiang University of Technology, Hangzhou 310023, China ' School of Cyberspace, Hangzhou Dianzi University, Hangzhou 310018, China ' School of Cyberspace, Hangzhou Dianzi University, Hangzhou 310018, China ' School of Cyberspace, Hangzhou Dianzi University, Hangzhou 310018, China
Abstract: Deep learning algorithm based on convolutional neural network has been widely used in the field of computer vision. A method based on deep convolution neural network is proposed for face recognition under low illumination. Firstly, the multi-scale retinex is used to enhance the face image in low-light imaging. Then the processed signal is input into the four-layer depth convolution neural network. The classification model is generated by the iterative training of the neural network. Finally, the input face image is classified based on the classification model. Multi-scale retinex utilises the principle of human eye perception of object brightness. Convolutional neural network can achieve better convergence rate and accuracy in classification and recognition of face images. Experiments on YaleB dataset show that the proposed algorithm and network model have better recognition performance.
Keywords: face recognition; deep learning; convolutional neural network; feature extraction; low illumination image; multi-scale retinex; face image enhancement.
DOI: 10.1504/IJAACS.2020.110746
International Journal of Autonomous and Adaptive Communications Systems, 2020 Vol.13 No.3, pp.260 - 272
Received: 28 Nov 2019
Accepted: 15 Feb 2020
Published online: 28 Oct 2020 *