We present a generic, efficient and iterative algorithm for interactively clustering classes of images and videos. The approach moves away from the use of large hand labelled tra...
We present a new object segmentation method that is based on geodesic active contours with combined shape and appearance priors. It is known that using shape priors can significan...
We describe an on-going project whose primary aim is to establish the technology of producing closed captions for TV news programs efficiently using natural language processing an...
Abstract--We describe some high-level approaches to estimating confidence scores for the words output by a speech recognizer. By "high-level" we mean that the proposed me...
We propose a new unsupervised learning technique for extracting information about authors and topics from large text collections. We model documents as if they were generated by a...
Michal Rosen-Zvi, Chaitanya Chemudugunta, Thomas L...