Algorithms for Detection, Tracking, Analysis, and Classification of Objects and Humans Actions
Stan Sclaroff
Boston University
Talk
Stan Sclaroff is a professor of computer science and the chair of the
Department of Computer Science at Boston University. He received the PhD
degree from the Massachusetts Institute of Technology in 1995.
In this talk, I will describe two projects in the Image and Video Computing research group at Boston University. In the first part of the talk, I will describe methods that learn a single family of detectors for object classes that exhibit large within-class variation. The effectiveness of this framework is demonstrated in experiments that involve detection, tracking, and pose estimation of human hands, faces, and vehicles in video. In the second part of the talk, I will describe methods for learning human action models from the Web. Our approach is unsupervised in the sense that it requires no human intervention other than the action keywords to be used to form text queries to Web image and video search engines. Thus, we can easily extend the vocabulary of actions, by simply making additional search engine queries. Experiments show the benefits of this approach in two areas: improving the retrieval precision of human action images, and tagging human actions in YouTube videos.
Collaborators in this work: Nazlı İkizler-Cinbiş, Shugao Ma, QuanYuan,
Vitaly Ablavsky, and Ashwin Thangali