This paper addresses the challenge of recognizing behavior of groups of individuals in unconstraint surveillance environments. As opposed to approaches that rely on agglomerative ...
Abstract. The prediction of diagnosis codes is typically based on freetext entries in clinical documents. Previous attempts to tackle this problem range from strictly rule-based sy...
Machine learning often relies on costly labeled data, and this impedes its application to new classification and information extraction problems. This has motivated the developme...
—This paper builds a generic modeling framework for analyzing the edge-creation process in dynamic random graphs in which nodes continuously alternate between active and inactive...
We address the problem of multi-person dataassociation-based tracking (DAT) in semi-crowded environments from a single camera. Existing trackletassociation-based methods using pur...