Topic: "Learning from competition"

Speaker: Weiguang Guan, SHARCNET

Zoom link

Generative Adversarial Networks (GANs) are one of the most interesting neural networks. The basic idea of GANs is learning a generative model through competition between two rivals. In this seminar we will demonstrate how to train a model to generate handwritten digits using GANs. We will explain how it works at code level. Some more advanced applications of GANs will be discussed at the end of the seminar and further reading materials will be provided for interested readers. Basic knowledge of Python and Tensorflow is preferred but not required.

Zoom account is required to attend the webinar. Need help attending a webinar? See the SHARCNET Help Wiki.