Local Convolutions Cause an Implicit Bias towards High Frequency Adversarial Examples

no code implementations • 19 Jun 2020 • Josue Ortega Caro, Yilong Ju, Ryan Pyle, Sourav Dey, Wieland Brendel, Fabio Anselmi, Ankit Patel

Inspired by theoretical work on linear full-width convolutional models, we hypothesize that the local (i. e. bounded-width) convolutional operations commonly used in current neural networks are implicitly biased to learn high frequency features, and that this is one of the root causes of high frequency adversarial examples.

Adversarial Robustness Vocal Bursts Intensity Prediction