Gabor Convolutional Networks
This work addresses the challenge of handling spatial transformations in object recognition for computer vision applications, representing an incremental improvement by integrating traditional filter properties into existing deep learning architectures.
The authors tackled the problem of deep convolutional neural networks lacking resistance to orientation and scale changes by incorporating Gabor filters into DCNNs, resulting in Gabor Convolutional Networks that achieve superior object recognition with fewer parameters and easier training.
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 end-to-end pipeline.