This paper presents an approach to unsupervised segmentation of moving and static objects occurring in a video. Objects are, in general, spatially cohesive and characterized by lo...
Standard cursive handwriting recognition is based on a language model, mostly a lexicon of possible word hypotheses or character n-grams. The result is a list of word alternatives...
Nowadays contents in Internet like weblogs, wikipedia and news sites become "live". How to notify and provide users with the relevant contents becomes a challenge. Unlike...
Weixiong Rao, Ada Wai-Chee Fu, Lei Chen 0002, Hanh...
In this paper, a new unsupervised hierarchical approach to textured color images segmentation is proposed. To this end, we have designed a two-step procedure based on a grey-scale...
This paper proposes a new tracking algorithm which combines object and background information, via building object and background appearance models simultaneously by nonparametric...