Sciweavers

1276 search results - page 103 / 256
» Statistical Models of Appearance for Computer Vision
Sort
View
ECCV
2000
Springer
16 years 5 months ago
A Probabilistic Background Model for Tracking
A new probabilistic background model based on a Hidden Markov Model is presented. The hidden states of the model enable discrimination between foreground, background and shadow. Th...
Jens Rittscher, Jien Kato, Sébastien Joga, ...
ICPR
2004
IEEE
16 years 4 months ago
To FRAME or not to FRAME in Probabilistic Texture Modelling?
The maximum entropy principle is a cornerstone of FRAME (Filters, RAndom fields, and Maximum Entropy) model considered at times as a first-ever step towards a universal theory of ...
Georgy L. Gimel'farb, Luc J. Van Gool, Alexey Zale...
BMVC
2001
15 years 5 months ago
Deformable Spatio-Temporal Shape Models: Extending ASM to 2D+Time
This paper extends 2D Active Shape Models to 2D+time by presenting a method for modelling and segmenting spatio-temporal shapes (ST-shapes). The modelling part consists of constru...
Ghassan Hamarneh, Tomas Gustavsson
ACCV
1998
Springer
15 years 7 months ago
Skin-Color Modeling and Adaptation
Abstract. This paper studies a statistical skin-color model and its adaptation. It is revealed that 1 human skin colors cluster in a small region in a color space; 2 the variance o...
Jie Yang, Weier Lu, Alex Waibel
ICPR
2010
IEEE
15 years 3 months ago
High-Level Feature Extraction Using SIFT GMMs and Audio Models
—We propose a statistical framework for high-level feature extraction that uses SIFT Gaussian mixture models (GMMs) and audio models. SIFT features were extracted from all the im...
Nakamasa Inoue, Tatsuhiko Saito, Koichi Shinoda, S...