June 9, 2023
11:00am - 12:00pm
Donald Bren Hall 6011

I will present the Segment Anything (SA) project: a new task, model, and dataset for image segmentation. Using our efficient model in an interactive data collection loop, we built the largest segmentation dataset to date, with over 1 billion masks on 11M licensed and privacy respecting images. The model is designed and trained to be promptable, so it can transfer zero-shot to a wide range of new image distributions and tasks with performance that is often competitive or superior to fully supervised approaches. The Segment Anything Model (SAM) has been released under an open Apache 2 license and the corresponding dataset (SA-1B) is also available to foster research into foundation models for computer vision.
Paper: https://arxiv.org/abs/2304.02643
Website: https://segment-anything.com

Alex Berg's research addresses computational visual recognition. He has worked on general object recognition and detection in images, action recognition in video, human pose identification in images, image parsing, face recognition, image search, combining natural language processing and computer vision, 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 UC Irvine. 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. Dr. Berg's work received the Marr Prize in 2013, the Everingham Prize for community contributions in 2016, the Helmholtz Prize for work that stood the test of time in 2017 & 2019, and the ACL Test of time award in 2022.