2784 search results - page 423 / 557 » Relation Extraction with Relation Topics |
ICCV
1998
IEEE
15 years 3 days ago
1998
IEEE
`Invariant regions' are image patches that automatically deform with changing viewpoint as to keep on covering identical physical parts of a scene. Such regions are then desc...
ECCV
2008
Springer
15 years 2 days ago
2008
Springer
Abstract. Statistical learning techniques have been used to dramatically speed-up keypoint matching by training a classifier to recognize a specific set of keypoints. However, the ...
ECCV
2008
Springer
15 years 2 days ago
2008
Springer
Abstract. The use of sparse invariant features to recognise classes of actions or objects has become common in the literature. However, features are often "engineered" to...
ECCV
2008
Springer
15 years 2 days ago
2008
Springer
Abstract. In this paper we present a new approach for establishing correspondences between sparse image features related by an unknown non-rigid mapping and corrupted by clutter an...
ICIP
2008
IEEE
14 years 12 months ago
2008
IEEE
This paper presents a distributed and occlusion-robust coding scheme for multi-view omnidirectional images, which relies on the geometry of the 3D scene. The Wyner-Ziv coder uses ...