less than 1 minute read

Introduction

Before Dall-E.2, Imagen, Stable Diffusion and all these magical AI artists came into this world, there was an algorithm (that is so intituitive and simple in it’s foundation) lying around that gives comparable performance, if not better than the general purpose LIMs and don’t need so much computation reseources.

GANs

Generative Adversial Networks, proposed by the team at University of Montreal (that included the AI pioneer and bookwriter Ian J. Goodfellow) ,

Updated: