DualNet: Learn Complementary Features for Image Recognition
Authors:
- Saihui HouSHSaihui Hou
- Xu LiuXLXu Liu
- Zilei WangZWZilei Wang
- 1 October 2017
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 502-510
Abstract
In this work we propose a novel framework named Dual-Net aiming at learning more accurate representation for image recognition. Here two parallel neural networks are coordinated to learn complementary features and thus a wider network is constructed. Specifically, we logically divide an end-to-end deep convolutional neural network into two functional parts, i.e., feature extractor and image classifier. The extractors of two subnetworks are placed side by side, which exactly form the feature extractor of DualNet. Then the two-stream features are aggregated to the final classifier for overall classification, while two auxiliary classifiers are appended behind the feature extractor of each subnetwork to make the separately learned features discriminative alone. The complementary constraint is imposed by weighting the three classifiers, which is indeed the key of DualNet. The corresponding training strategy is also proposed, consisting of iterative training and joint fine tuning, to make the two subnetworks cooperate well with each other. Finally, DualNet based on the well-known CaffeNet, VGGNet, NIN and ResNet are thoroughly investigated and experimentally evaluated on multiple datasets including CIFAR-100, Stanford Dogs and UEC FOOD-100. The results demonstrate that DualNet can really help learn more accurate image representation, and thus result in higher accuracy for recognition. In particular, the performance on CIFAR-100 is state-of-the-art compared to the recent works.Keywords
This publication has 13 references indexed in Scilit:
- Deep Decision Network for Multi-class Image Classification2016
- Deep Residual Learning for Image Recognition2016
- Attention to Scale: Scale-Aware Semantic Image Segmentation2016
- Wide Residual Networks2016
- Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classification2015
- HD-CNN: Hierarchical Deep Convolutional Neural Networks for Large Scale Visual Recognition2015
- Going deeper with convolutions2015
- Histograms of Oriented Gradients for Human Detection2005
- Gradient-based learning applied to document recognitionProceedings of the IEEE, 1998
- Neocognitron: A self-organizing neural network model for a mechanism of pattern recognition unaffected by shift in positionBiological Cybernetics, 1980