Abstract:Though widely used in image classification, convolutional neural networks (CNNs) are prone to noise interruptions, i.e. the CNN output can be drastically changed by small image noise. To improve the noise robustness, we try to integrate CNNs with wavelet by replacing the common down-sampling (max-pooling, strided-convolution, and average pooling) with discrete wavelet transform (DWT). We firstly propose general DWT and inverse DWT (IDWT) layers applicable to various orthogonal and biorthogonal discrete wavelets like Haar, Daubechies, and Cohen, etc., and then design wavelet integrated CNNs (WaveCNets) by integrating DWT into the commonly used CNNs (VGG, ResNets, and DenseNet). During the down-sampling, WaveCNets apply DWT to decompose the feature maps into the low-frequency and high-frequency components. Containing the main information including the basic object structures, the low-frequency component is transmitted into the following layers to generate robust high-level features. The high-frequency components are dropped to remove most of the data noises. The experimental results show that WaveCNets achieve higher accuracy on ImageNet than various vanilla CNNs. We have also tested the performance of WaveCNets on the noisy version of ImageNet, ImageNet-C and six adversarial attacks, the results suggest that the proposed DWT/IDWT layers could provide better noise-robustness and adversarial robustness. When applying WaveCNets as backbones, the performance of object detectors (i.e., faster R-CNN and RetinaNet) on COCO detection dataset are consistently improved. We believe that suppression of aliasing effect, i.e. separation of low frequency and high frequency information, is the main advantages of our approach. The code of our DWT/IDWT layer and different WaveCNets are available at https://github.com/CVI-SZU/WaveCNet.

A Down-sampling Method Based on the Discrete Wavelet Transform for CNN Classification

Research on Image Classification Method of Features of Combinatorial Convolution

Image Classification Using Convolutional Neural Network with Wavelet Domain Inputs

WaveCNet: Wavelet Integrated CNNs to Suppress Aliasing Effect for Noise-Robust Image Classification

A Pruning Method Based on the Dissimilarity of Angle among Channels and Filters

Cascade wavelet transform based convolutional neural networks with application to image classification

Dilated Residual Encode-Decode Networks for Image Denoising

Learning Efficient Convolutional Networks Through Network Slimming.

Volume Upscaling with Convolutional Neural Networks

Wavelet Integrated CNNs for Noise-Robust Image Classification

Mixed-Precision Quantization for CNN-Based Remote Sensing Scene Classification

Automatic channel pruning via clustering and swarm intelligence optimization for CNN

Coarse-to-fine Trained Multi-Scale Convolutional Neural Networks for Image Classification.

SCWC: Structured channel weight sharing to compress convolutional neural networks

WaveCNNs-AT: Wavelet-based deep CNNs of adaptive threshold for signal recognition

Lightweight Deep Neural Network Model With Padding-Free Downsampling

ChannelNets: Compact and Efficient Convolutional Neural Networks via Channel-Wise Convolutions

Deep Sampling Networks

Research on improved wavelet convolutional wavelet neural networks

Compressing Deep Convolutional Networks using Vector Quantization

Low-res MobileNet: An efficient lightweight network for low-resolution image classification in resource-constrained scenarios