IEEE Chicago's Information Theory Society Webinar: "Deep Generative Models and Inverse Problems"

#machine #learning; #deep #generative #models; #inverse #problems; #generalizing #compressed #sensing; #information #theory
Share

Abstract:  Modern deep generative models like GANs, VAEs and invertible flows are achieving amazing results on modeling high-dimensional distributions, especially for images. We will show how they can be used to solve inverse problems by generalizing compressed sensing beyond sparsity, extending the theory of Restricted Isometries to sets created by generative models. We will present the general framework, new results and open problems in this space.

Bio: Alex Dimakis is a Professor at the Electrical and Computer Engineering department, University of Texas at Austin. He received his Ph.D. from UC Berkeley and the Diploma degree from the National Technical University of Athens. He received several awards including the IEEE Information Theory Society James Massey Award, NSF Career, a Google research award, and the Eli Jury dissertation award and the joint IEEE Information Theory and Communications Society Best Paper Award. His research interests include information theory, coding theory and machine learning with a current focus on unsupervised learning.  


  Date and Time

  Location

  Hosts

  Registration


  • Date: 04 Dec 2020
  • Time: 11:00 AM to 12:15 PM
  • All times are (UTC-06:00) Central Time (US & Canada)
  • Add_To_Calendar_icon Add Event to Calendar
If you are not a robot, please complete the ReCAPTCHA to display virtual attendance info.
  • Chicago, Illinois
  • United States 60607
  • Contact Event Host
  • Co-sponsored by University of Illinois Chicago -- Electrical and Computer Engineering Department
  • Starts 10 November 2020 09:02 PM
  • Ends 03 December 2020 11:55 PM
  • All times are (UTC-06:00) Central Time (US & Canada)
  • No Admission Charge

  Speakers

Dr Dimakis

Topic:

Talk: Deep Generative models and Inverse Problems

Abstract:  Modern deep generative models like GANs, VAEs and invertible flows are achieving amazing results on modeling high-dimensional distributions, especially for images. We will show how they can be used to solve inverse problems by generalizing compressed sensing beyond sparsity, extending the theory of Restricted Isometries to sets created by generative models. We will present the general framework, new results and open problems in this space.

Biography:

Bio: Alex Dimakis is a Professor at the Electrical and Computer Engineering department, University of Texas at Austin. He received his Ph.D. from UC Berkeley and the Diploma degree from the National Technical University of Athens. He received several awards including the IEEE Information Theory Society James Massey Award, NSF Career, a Google research award, and the Eli Jury dissertation award and the joint IEEE Information Theory and Communications Society Best Paper Award. His research interests include information theory, coding theory and machine learning with a current focus on unsupervised learning.  More info at https://users.ece.utexas.edu/~dimakis/.