An Effective Anti-Aliasing Approach for Residual Networks

Abstract

Image pre-processing in the frequency domain has traditionally played a vital role in computer vision and was even part of the standard pipeline in the early days of deep learning. However, with the advent of large datasets, many practitioners concluded that this was unnecessary due to the belief that these priors can be learned from the data itself. Frequency aliasing is a phenomenon that may occur when sub-sampling any signal, such as an image or feature map, causing distortion in the sub-sampled output. We show that we can mitigate this effect by placing non-trainable blur filters and using smooth activation functions at key locations, particularly where networks lack the capacity to learn them. These simple architectural changes lead to substantial improvements in out-of-distribution generalization on both image classification under natural corruptions on ImageNet-C [10] and few-shot learning on Meta-Dataset [17], without introducing additional trainable parameters and using the default hyper-parameters of open source codebases.

References in corpus (3)

• Accurate, Large Minibatch SGD: Training ImageNet in 1 Hour
• AutoAugment: Learning Augmentation Policies from Data
• Smooth Adversarial Training

Cited by in corpus (8)

• MedViT: A Robust Vision Transformer for Generalized Medical Image Classification
• Towards Robust Vision Transformer
• Alias-Free Convnets: Fractional Shift Invariance via Polynomial Activations
• Improving Sound Event Classification by Increasing Shift Invariance in Convolutional Neural Networks
• Bridging the Gap Between Adversarial Robustness and Optimization Bias
• Using Synthetic Corruptions to Measure Robustness to Natural Distribution Shifts
• Combining Different V1 Brain Model Variants to Improve Robustness to Image Corruptions in CNNs
• Training Efficiency and Robustness in Deep Learning