Lecture: Evaluating Generative Adversarial Networks for Medical Image Synthesis and Restoration by Prasanth Ganesan on 11/20/18

Evaluating Generative Adversarial Networks for Medical Image Synthesis and Restoration

Brown Bag Lecture by Prasanth Ganesan | 11/20/2018 11:00AM – 12PM | 7th Floor Conference Room, Bldg 38A

Abstract:

Generative adversarial networks (GANs) have become widely popular due to its ability to learn a data distribution in an unsupervised manner without the use of computationally expensive Markov Chains. Today, GANs are being modelled for various tasks such as generating fake images – E.g., fake images of X-rays and MRIs; translating images from one modality to another – E.g., a photograph to art, a CT to an MRI; and restoring corrupted images – E.g., removal of camera blur, increasing the resolution of poor-quality images. This talk will provide an overview of the most popular models of GANs and their applications. In particular, Deep Convolutional GAN (DC-GAN), Self-Attention GAN (SA-GAN) and Progressive Growing GAN (PG-GAN) will be discussed. The talk will focus on GANs implemented at CEB for X-rays and colposcopy  images.

Bio:

Prasanth is a fourth year PhD candidate in Electrical Engineering at Florida Atlantic University. He is currently a Graduate Research Fellow at NLM working under Dr. Sameer Antani, on Generative Adversarial Networks. He earned his Masters in Electrical Engineering from Rochester Institute of Technology in 2015, and his Bachelor’s degree in Electronics Engineering from Anna University, India, in 2013. Prasanth’s research expertise is in developing signal processing and pattern recognition algorithms. He aspires to work on innovative bioengineering methods in the future.

Lecture: Cervical Cytology Image Analysis and Classification Using Graph-Based Techniques and Deep Learning by Sudhir Sornapudi on 11/27/18
Lecture: Integrative Sparse Modeling and Classification of Biomedical Imaging Patterns by Dr. Keni Zheng on 11/6/18