700 search results - page 102 / 140 » Face Recognition from Video: A CONDENSATION Approach |
CAIP
2007
Springer
15 years 3 months ago
2007
Springer
In this paper, we propose a new manifold representation capable of being applied for visual speech recognition. In this regard, the real time input video data is compressed using P...
123
click to vote
KDD
2001
ACM
15 years 2 months ago
2001
ACM
This paper describes the Perseus project, which is devoted to developing techniques and tools for creating personalized multimedia news portals. The purpose of a personalized mult...
CVPR
2007
IEEE
15 years 4 months ago
2007
IEEE
3D human pose recovery is considered as a fundamental step in view-invariant human action recognition. However, inferring 3D poses from a single view usually is slow due to the la...
CVPR
2010
IEEE
2010
IEEE
Learning a Hierarchy of Discriminative Space-Time Neighborhood Features for Human Action Recognition
15 years 6 months ago
Recent work shows how to use local spatio-temporal features to learn models of realistic human actions from video. However, existing methods typically rely on a predefined spatial...
SCIA
2007
Springer
15 years 3 months ago
2007
Springer
Computing high quality 3D models from multi-view stereo reconstruction is an active topic as can be seen in a recent review [15]. Most approaches make the strong assumption that th...