Mixture of deep CNN-based ensemble model for image retrieval
Hsin-Kai Huang, Chien-Fang Chiu, Chien-Hao Kuo, Yu‐Chi Wu, Narisa Nan Chu, Pao‐Chi Chang · 2016
This paper proposes an aggregate (or mixture) of ensemble models for image retrieval based on deep Convolutional Neural Networks (CNN). It utilizes two kinds of deep learning networks, AlexNet and Network In Network (NIN), to obtain image features, and to compute weighted average feature vectors for image retrieval. Based on experimental results, the aggregate ensemble architecture effectively enhances learning with higher accuracy than single CNN in image classification. When the proposed aggregate of deep CNN-based ensemble model is applied to CIFAR-10 and CIFAR-100 datasets, it is shown to achieve 0.867 and 0.526 mean average precision in image retrieval, respectively.