Downscaling Attack and Defense: Turning What You See Back Into What You Get

Abstract

The resizing of images, which is typically a required part of preprocessing for computer vision systems, is vulnerable to attack. Images can be created such that the image is completely different at machine-vision scales than at other scales and the default settings for some common computer vision and machine learning systems are vulnerable. We show that defenses exist and are trivial to administer provided that defenders are aware of the threat. These attacks and defenses help to establish the role of input sanitization in machine learning.

References in corpus (9)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• Feature Squeezing: Detecting Adversarial Examples in Deep Neural Networks
• Targeted Backdoor Attacks on Deep Learning Systems Using Data Poisoning
• Towards Deep Neural Network Architectures Robust to Adversarial Examples
• Rethinking the Inception Architecture for Computer Vision
• Certified Adversarial Robustness with Additive Noise
• Motivating the Rules of the Game for Adversarial Example Research
• Delving Deeper into Anti-aliasing in ConvNets
• Wolf in Sheep's Clothing - The Downscaling Attack Against Deep Learning Applications