Hacking Generative Models with Differentiable Network Bending

Abstract

In this work, we propose a method to ‘hack’ generative models, pushing their outputs away from the original training distribution towards a new objective. We inject a small-scale trainable module between the intermediate layers of the model and train it for a low number of iterations, keeping the rest of the network frozen. The resulting output images display an uncanny quality, given by the tension between the original and new objectives that can be exploited for artistic purposes.

Publication
NeurIPS 2023, ML for Creativity and Design
Ahmed Youssef
Ahmed Youssef
Ph.D. Candidate

My research interests include generative modeles, probalistic modeling, and there application in science and outside.