Fereshteh Sadeghi


Hometown: Tehran, Iran
University of Washington

Student Page: http://homes.cs.washington.edu/~fsadeghi

Research: Fereshteh's research is focused towards making autonomous robot controllers using deep reinforcement learning and vision. Her goal is to take advantage of advances in 3D graphics simulations and deep learning to train robust and generalizable controllers that can be transferred to the real world. In the past, she has worked on various problems in visual understanding and recognition. 

Bio: Fereshteh Sadeghi is a Computer Science PhD candidate at University of Washington and a visiting student at University of California, Berkeley. She is a member of Berkeley AI Research (BAIR) Lab as well as Graphics and Imaging Laboratory (GRAIL) at UW. Fereshteh is advised by Prof. Sergey Levine and her area of research is at the intersection of robotics, computer vision and machine learning. Her focus is on developing new deep reinforcement learning methods to learn robot controllers with high generalization capability that can perform tasks in diverse real-world settings.