Adversarial Noise Workshop

A course on how to modify the input data on a neural not to lead it to false predictions.

Based on the workshop notes of Javier Lloret Pardo I created a Jupyter Book to turn the material into an interactive website.

The book is avalibale at https://capital-g.github.io/adversarial-noise-workshop/.

The source code of the book is hoste at GitHub.

Image of a harmonica, heavily modified by adversarial noise.