Impact of Aliasing on Generalization in Deep Convolutional Networks

by   Cristina Vasconcelos, et al.

We investigate the impact of aliasing on generalization in Deep Convolutional Networks and show that data augmentation schemes alone are unable to prevent it due to structural limitations in widely used architectures. Drawing insights from frequency analysis theory, we take a closer look at ResNet and EfficientNet architectures and review the trade-off between aliasing and information loss in each of their major components. We show how to mitigate aliasing by inserting non-trainable low-pass filters at key locations, particularly where networks lack the capacity to learn them. These simple architectural changes lead to substantial improvements in generalization on i.i.d. and even more on out-of-distribution conditions, such as image classification under natural corruptions on ImageNet-C [11] and few-shot learning on Meta-Dataset [26]. State-of-the art results are achieved on both datasets without introducing additional trainable parameters and using the default hyper-parameters of open source codebases.


page 1

page 7

page 14

page 16


Generalization to translation shifts: a study in architectures and augmentations

We provide a detailed evaluation of various image classification archite...

On the Blindspots of Convolutional Networks

Deep convolutional network has been the state-of-the-art approach for a ...

Understanding Deep Convolutional Networks

Deep convolutional networks provide state of the art classifications and...

Hyper-parameter optimization of Deep Convolutional Networks for object recognition

Recently sequential model based optimization (SMBO) has emerged as a pro...

Using mixup as regularization and tuning hyper-parameters for ResNets

While novel computer vision architectures are gaining traction, the impa...

Parametric Variational Linear Units (PVLUs) in Deep Convolutional Networks

The Rectified Linear Unit is currently a state-of-the-art activation fun...

Towards Principled Design of Deep Convolutional Networks: Introducing SimpNet

Major winning Convolutional Neural Networks (CNNs), such as VGGNet, ResN...