In this paper, we present a systematic framework for re-cognizing realistic actions from videos “in the wild.” Such unconstrained videos are abundant in personal collections as...
Jingen Liu (University of Central Florida), Jiebo ...
An unsupervised clustering of the webpages on a website is a primary requirement for most wrapper induction and automated data extraction methods. Since page content can vary dras...
In this paper, we exploit the problem of inferring images’ semantic concepts from community-contributed images and their associated noisy tags. To infer the concepts more accura...