| | |
Summary: Learning Bayesian Classifiers for a Visual Grammar
Selim Aksoy, Krzysztof Koperski, Carsten Tusk, Giovanni Marchisio
Insightful Corporation
1700 Westlake Ave. N., Suite 500
Seattle, WA 98109, USA
{saksoy,krisk,ctusk,giovanni}@insightful.com
James C. Tilton
NASA Goddard Space Flight Center
Mail Code 935
Greenbelt, MD 20771, USA
James.C.Tilton@nasa.gov
Abstract-- A challenging problem in image content extraction
and classification is building a system that automatically learns
high-level semantic interpretations of images. We describe a
Bayesian framework for a visual grammar that aims to reduce
the gap between low-level features and user semantics. Our
approach includes learning prototypes of regions and their spatial
relationships for scene classification. First, naive Bayes classifiers
perform automatic fusion of features and learn models for
region segmentation and classification using positive and negative
|
