In this event, we are thrilled to host Dr. Sina Honari a postdoctoral researcher from the Computer Vision Lab at EPFL who will talk about: Learning more efficiently human poses with inductive priors. Abstract 3D human pose estimation has applications in action recognition, motion analysis, human tracking and representation. For example, it can be used in sports events such as athletics or diving to help judges score the athlete's performance or to help coaches analyze the movement of the athletes. Another application is in augmented and virtual reality where the estimated pose can be used to represent human avatars. In this talk, Dr. Sina Honari will present different ways of injecting inductive priors into the learning framework of neural networks, where the goal is to facilitate information extraction or representation given the dynamics of the input data or the task of interest. While some of the proposed approaches are more general, they are evaluated on human pose estimation or mesh recovery. PS: To stay up to date with events, become part of our GDSC community (700+ members) by clicking the 'Join us' button here!
CVlab @ EPFL
postdoctoral researcher
CVLab @ EPFL
Computer Vision Research Engineer
ETH Zurich
GDSC Lead
ETH Zürich
GDSC Lead
ETH Zürich
GDSC Lead
Core Team Member
ETH Zurich
Core Team Member
ETH Zurich
Core Team Member
Core Team Member
Core Team Member
Core Team Member
ETH
Core Team Member
Core Team Member
Core Team Member
Core Team Member
Core Team Member
ETH Zürich
Core Team Member
Core Team Member