Ben Taskar
Computer Science Department
University of Pennsylvania
"From Co-occurrence to Correspondence"
Abstract:
While supervised learning methods for classification and structured prediction are very effective in many domains, they require detailed and precise labeling of large amounts of data. Weakly or ambiguously labeled data, presents major challenges as well as opportunities. For example, to build a machine translation system, we typically have large amounts of translated sentences to learn from, but without word or phrase level correspondence. Copious images and videos on the web or your hard drive are typically labeled with captions of who and what is in the picture, but not where and when. The challenges are both theoretical and algorithmic: under what assumptions can we guarantee effective and efficient learning of precise correspondence from pure co-occurence? I will describe our ongoing work on weakly supervised learning approaches for machine translation and parsing of images, videos and text.
Tuesday, November 11, 2008
3:00 - 4:15
Wu & Chen
101 Levine Hall