114 search results - page 20 / 23 » Learning Appearance Manifolds from Video |
195
click to vote
Publication
State-of-the-art person re-identi cation methods seek robust person matching through combining various feature types. Often, these features are implicitly assigned with a single ve...
FGR
2011
IEEE
12 years 11 months ago
2011
IEEE
Abstract— We propose here to acquire high resolution sequences of a person’s face using a pan-tilt-zoom (PTZ) network camera. This capability should prove helpful in forensic a...
BMVC
2010
13 years 5 months ago
2010
Off-line trained class-specific object detectors are designed to detect any instance of the class in a given image or video sequence. In the context of object tracking, however, o...
ICCV
2005
IEEE
14 years 1 months ago
2005
IEEE
We address the problem of learning object class models and object segmentations from unannotated images. We introduce LOCUS (Learning Object Classes with Unsupervised Segmentation...
ICIP
2007
IEEE
13 years 11 months ago
2007
IEEE
The major problem in building a good lipreading system is to extract effective visual features from enormous quantity of video sequences data. For appearance-based feature analysi...