Gabor Convolutional Networks

Abstract

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.

References in corpus (9)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• Improving neural networks by preventing co-adaptation of feature detectors
• ADADELTA: An Adaptive Learning Rate Method
• Deformable Convolutional Networks
• Gabor Convolutional Networks
• Gabor Filter Assisted Energy Efficient Fast Learning Convolutional Neural Networks
• One-Two-One Networks for Compression Artifacts Reduction in Remote Sensing
• Latent Constrained Correlation Filter
• Oriented Response Networks

Cited by in corpus (38)

• Object Detection in Optical Remote Sensing Images: A Survey and A New Benchmark
• Gabor Convolutional Networks
• NAS-FAS: Static-Dynamic Central Difference Network Search for Face Anti-Spoofing
• Fully Learnable Deep Wavelet Transform for Unsupervised Monitoring of High-Frequency Time Series
• What Do We Understand About Convolutional Networks?
• Deep Retinal Image Segmentation with Regularization Under Geometric Priors
• Local Intensity Order Transformation for Robust Curvilinear Object Segmentation
• Lightweight Pixel Difference Networks for Efficient Visual Representation Learning
• Searching Central Difference Convolutional Networks for Face Anti-Spoofing
• Scale-covariant and scale-invariant Gaussian derivative networks
• Learning the Pareto Front with Hypernetworks
• Resolution learning in deep convolutional networks using scale-space theory
• Provably scale-covariant continuous hierarchical networks based on scale-normalized differential expressions coupled in cascade
• GaborNet: Gabor filters with learnable parameters in deep convolutional neural networks
• Pixel Difference Networks for Efficient Edge Detection
• Multi-Modal Face Anti-Spoofing Based on Central Difference Networks
• Some open questions on morphological operators and representations in the deep learning era
• Towards Learning Affine-Invariant Representations via Data-Efficient CNNs
• Interpretable Deep Feature Propagation for Early Action Recognition
• 3D Segmentation with Fully Trainable Gabor Kernels and Pearson's Correlation Coefficient
• Learning Complex Basis Functions for Invariant Representations of Audio
• On the Realization and Analysis of Circular Harmonic Transforms for Feature Detection
• Deformable Gabor Feature Networks for Biomedical Image Classification
• Learnable Gabor modulated complex-valued networks for orientation robustness
• ExplainFix: Explainable Spatially Fixed Deep Networks
• Rega-Net:Retina Gabor Attention for Deep Convolutional Neural Networks
• Naive Gabor Networks for Hyperspectral Image Classification
• Scale-Invariant Multi-Oriented Text Detection in Wild Scene Images
• Convolutional Neural Networks: A Binocular Vision Perspective
• Star DGT: a Robust Gabor Transform for Speech Denoising
• Time-Frequency Localization Using Deep Convolutional Maxout Neural Network in Persian Speech Recognition
• A Local Descriptor with Physiological Characteristic for Finger Vein Recognition
• Multifaceted Analysis of Fine-Tuning in Deep Model for Visual Recognition
• Filter Bank Regularization of Convolutional Neural Networks
• PCNet: A Structure Similarity Enhancement Method for Multispectral and Multimodal Image Registration
• A Cascaded Zoom-In Network for Patterned Fabric Defect Detection
• Orientation Convolutional Networks for Image Recognition
• Gabor filter incorporated CNN for compression