Gabor Convolutional Networks

作者：Shangzhen Luan, Chen Chen, Baochang Zhang, J. Han, Jianzhuang Liu · 发表于：IEEE Workshop/Winter Conference on Applications of Computer Vision · 年份：2017 · DOI：10.1109/TIP.2018.2835143 · 被引用次数：400 · 研究领域：Medicine、Computer Science、Mathematics

Steerable properties dominate the design of traditional filters, e.g., Gabor filters, and endow features the capability of dealing with spatial transformations. However, such excellent properties have not been well explored in the popular deep convolutional neural networks (DCNNs). In this paper, we propose a new deep model, termed Gabor Convolutional Networks (GCNs or Gabor CNNs), which incorporates Gabor filters into DCNNs to enhance the resistance of deep learned features to the orientation and scale changes. By only manipulating the basic element of DCNNs based on Gabor filters, i.e., the convolution operator, GCNs can be easily implemented and are compatible with any popular deep learning architecture. Experimental results demonstrate the super capability of our algorithm in recognizing objects, where the scale and rotation changes occur frequently. The proposed GCNs have much fewer learnable network parameters, and thus is easier to train with an endtoend pipeline. The source code will be here 1.