Speaker: Alex Berg
Affiliation: University of North Carolina at Chapel Hill
ABSTRACT: Vision is an amazingly useful way to sense the world around us and poses immense computational challenges or opportunities, depending on one’s perspective. Vision tasks involve both the complexities of physical structure and nuances of how humans (or robots!) communicate and reason about that structure. While these tasks have remained consistent, the history of research in computer vision has been very much both driven and tied to what computation is available. I will talk about some of our work pushing the boundaries of what could be recognized in visual data using computers, including the ImageNet Challenge and connecting natural language and computer vision. This push resulted in advances in algorithms, including some of the more recent work from my group on very efficient object detection. This work addresses datasets, algorithms, and computation down to FPGAs, and is currently exploring how to integrate our success in deep learning into vision approaches that can be used in everyday robotics. BIO: Alex Berg’s research concerns computational visual recognition. He has worked on general object recognition in images, action recognition in video, human pose identification in images, image parsing, face recognition, image search, and machine learning for computer and human vision. He co-founded and co-organized the ImageNet Large Scale Visual Recognition Challenge, and organized the first Large-Scale Learning for Vision workshop. He is currently an associate professor in computer science at UNC Chapel Hill. Prior to that he was on the faculty at Stony Brook University, a research scientist at Columbia University, and research scientist at Yahoo! Research. His PhD at U.C. Berkeley developed a novel approach to deformable template matching. He earned a BA and MA in Mathematics from Johns Hopkins University and learned to race sailboats at SSA in Annapolis. His work received the Marr Prize in 2013, the Everingham Prize for community contributions in 2016, and the Helmholtz Prize for work that stood the test of time in 2017.
Hosted by Professor Stefano Soatto
Date(s) - Feb 15, 2018
4:15 pm - 5:45 pm