10 search results - page 1 / 2 » Mutual Boosting for Contextual Inference |
NIPS
2003
13 years 6 months ago
2003
Mutual Boosting is a method aimed at incorporating contextual information to augment object detection. When multiple detectors of objects and parts are trained in parallel using A...
CVPR
2005
IEEE
14 years 6 months ago
2005
IEEE
We present a novel approach for fast object class recognition incorporating contextual information into boosting. The object is represented as a constellation of generalized corre...
MIR
2004
ACM
13 years 10 months ago
2004
ACM
We present a new framework for characterizing and retrieving objects in cluttered scenes. This CBIR system is based on a new representation describing every object taking into acc...
NIPS
2004
13 years 6 months ago
2004
We seek to both detect and segment objects in images. To exploit both local image data as well as contextual information, we introduce Boosted Random Fields (BRFs), which use boos...
DLOG
2007
13 years 7 months ago
2007
Abstract. In the paper we propose a method of structuring a knowledge base into hierarchically related contexts and present how this arrangement influences the structure of TBox a...