This paper describes a series of experiments to test the hypothesis that the parallel application of multiple NLP tools and the integration of their results improves the correctne...
Retrieving human actions from video databases is a paramount but challenging task in computer vision. In this work, we develop such a framework for robustly recognizing human acti...
In this paper we compare the performance of tree different dynamic traffic engineering algorithms exploiting path diversity in the Internet, TEXCP, TRUMP and MIRTO. We passed throu...
Specular flow is an important class of optical flow whose utility in visual tasks has gained much interest in contemporary vision research. Unfortunately, however, reliably estima...