Deep Learning @ UGA

Generative Adversarial Networks (GANs)

Generative Adversarial Networks (GANs) are an exciting new deep learning technique pioneered by Ian Goodfellow and his research lab. This new model describes a framework for training competing models that perform opposite tasks: generating data and discriminating the veracity of training instances. By training a generative and discriminative model we can tackle a range of tasks from generating more accurate representations of real data as well as improving semi-supervised learning techniques.

Slides [pdf]

Rajeswari Sivakumar is a 2nd year graduate student at the University of Georgia, studying Artificial Intelligence. Her research interests include applications of machine learning in biomedical image analysis. She is currently examining deep learning approaches to classifying brain images.