Multi-layer CNN Features Fusion and Classifier Optimization for Face Recognition
Yulin Wu, Mingyan Jiang · Proceedings of the 2018 2nd International Conference on Computer Science and Artificial Intelligence · 2018
Convolutional neural network (CNN) is widely used in face recognition. However, the fully connected layer features of CNN are difficult to express the facial information completely and the structure of CNN also lacks a proper classifier. In this paper, a model based on multi-layer CNN features is proposed, which contains optimized classifiers. The features of all convolutional layers and fully connected layer are extracted to enhance image representation. Moreover, the model adopts the support vector machine (SVM) optimized by artificial bee colony (ABC) algorithm as the optimal classifier. The results are obtained by fusing the outputs of classifiers. Experimental results illustrate that the proposed method outperforms the compared state-of-art algorithms with several experiments operating in the face databases of ORL and FERET.