This paper presents a novel approach to pedestrian classification which involves utilizing the synthesized virtual samples of a learned generative model to enhance the classificat...
The state of the art in visual object retrieval from large databases is achieved by systems that are inspired by text retrieval. A key component of these approaches is that local ...
James Philbin, Ondrej Chum, Michael Isard, Josef S...
In this paper, a new approach for object detection and pose estimation is introduced. The contribution consists in the conception of entities permitting stable detection and relia...
Stefan Hinterstoisser, Selim Benhimane, Nassir Nav...
In this paper we propose a novel inhomogeneous Gibbs model by the minimax entropy principle, and apply it to face modeling. The maximum entropy principle generalizes the statistic...
We propose to incorporate a priori geometric constraints in a 3?D stereo reconstruction scheme to cope with the many cases where image information alone is not sufficient to accur...