Vision and Autonomous Systems Seminar

  • Gates Hillman Centers
  • Traffic21 Classroom 6501

Geometric Deep Learning for Perceiving and Modeling Humans

Perceiving and modeling shape and appearance of the human body from single images is a severely under-constrained problem that not only requires large volumes of data, but also prior knowledge.  In this talk I will present recent solutions on how deep learning can leverage on geometric reasoning to address tasks like 3D estimation of the body pose and the shape of the clothes. In the second part of the talk I will also describe novel GAN’s conditioning schemes to model full body’s appearance and facial expressions in a weakly supervised manner.

Francesc Moreno Noguer received the MSc degrees in industrial engineering and electronics from the Technical University of Catalonia (UPC) and the Universitat de Barcelona in 2001 and 2002, respectively, and the PhD degree from UPC in 2005. From 2006 to 2008, he was a postdoctoral fellow at the computer vision departments of Columbia University and the Ecole Polytecnique Fédéralede Lausanne. In 2009, he joined the Institutde Robòtica I Informàtica Industrial in Barcelona as an associate researcher of the Spanish Scientific Research Council. His research interests are in Computer Vision and Machine Learning, with topics including human shape and motion estimation, 3D reconstruction of rigid and nonrigid shape and camera pose estimation from single images and video sequences. He received best paper honorable mention award at ECCV'18, best paper awards at ICCV workshop on Fashion'17, Machine Vision Applications'15, JornadasAutomática'14 and Ibpria'05, UPC’s Doctoral Dissertation Extraordinary Award in 2008, outstanding reviewer awards at ECCV'12 and CVPR'14, and Amazon research award (2019) and Google Faculty research awards (2017, 2019).

Refreshments Served.

Sponsored in part by Facebook Reality Labs Pittsburgh

For More Information, Please Contact: