Time for a Background Check! Uncovering the impact of Background Features on Deep Neural Networks

by   Vikash Sehwag, et al.
Princeton University

With increasing expressive power, deep neural networks have significantly improved the state-of-the-art on image classification datasets, such as ImageNet. In this paper, we investigate to what extent the increasing performance of deep neural networks is impacted by background features? In particular, we focus on background invariance, i.e., accuracy unaffected by switching background features and background influence, i.e., predictive power of background features itself when foreground is masked. We perform experiments with 32 different neural networks ranging from small-size networks to large-scale networks trained with up to one Billion images. Our investigations reveal that increasing expressive power of DNNs leads to higher influence of background features, while simultaneously, increases their ability to make the correct prediction when background features are removed or replaced with a randomly selected texture-based background.


page 2

page 4


Deep Neural Network Concepts for Background Subtraction: A Systematic Review and Comparative Evaluation

Conventional neural networks show a powerful framework for background su...

Impact of ImageNet Model Selection on Domain Adaptation

Deep neural networks are widely used in image classification problems. H...

Scene Image Representation by Foreground, Background and Hybrid Features

Previous methods for representing scene images based on deep learning pr...

Background Hardly Matters: Understanding Personality Attribution in Deep Residual Networks

Perceived personality traits attributed to an individual do not have to ...

Training BatchNorm and Only BatchNorm: On the Expressive Power of Random Features in CNNs

Batch normalization (BatchNorm) has become an indispensable tool for tra...

Survey of Expressivity in Deep Neural Networks

We survey results on neural network expressivity described in "On the Ex...

Please sign up or login with your details

Forgot password? Click here to reset