Large margin classifiers have demonstrated their advantages in many visual learning tasks, and have attracted much attention in vision and image processing communities. In this pa...
We present a method which enables rapid and dense reconstruction of scenes browsed by a single live camera. We take point-based real-time structure from motion (SFM) as our starti...
Two-tone (“Mooney”) images seem to arouse vivid 3D percept of faces, both familiar and unfamiliar, despite their seemingly poor content. Recent psychological and fMRI studies ...
Ira Kemelmacher-Shlizerman, Ronen Basri, Boaz Nadl...
This paper introduces a novel analogy with the way in which honeybee colonies operate in order to solve the problem of sparse and quasi dense reconstruction. To successfully solve...
In this paper we present a system for classifying various human actions in compressed domain video framework. We introduce the notion of quantifying the motion involved, through w...