Generative Adversarial Networks Discussion



I was enrolled in a seminar course titled Learning with Limited Supervision in Fall 2019, and each class a group of students would lead a discussion on a chosen paper. For my discussion, I chose the Generative Adversarial Networks by Ian Goodfellow.

I was tasked with not only presenting the methodology, but also debating the strengths of the paper against my colleague who presented the paper’s weaknesses.

Despite the paper being a seminal work in Computer Vision and in Deep Learning, there was plenty to discuss about the ideas presented in this paper since a few years have passed since its publishing, which allows us to look back and reflect on how these ideas have held up over the years.

Muhammad "Osama" Sakhi
Head Teaching Assistant for Introduction to Artificial Intelligence

MS CSE student at GT