PASCAL - Pattern Analysis, Statistical Modelling and Computational Learning

Improving Kernel Classifiers for Object Categorization Problems
Alexei Pozdnoukov and Samy Bengio
In: Learning with Partly Classified Training Data (ICML'05 workshop), 7 Aug 2005, Bohn, Germany.

Abstract

This paper presents an approach for improving the performance of kernel classifiers applied to object categorization problems. The approach is based on the use of distributions centered around each training points, which are exploited for inter-class invariant image representation with local invariant features. Furthermore, we propose an extensive use of unlabeled images for improving the SVM-based classifier.

PDF - Requires Adobe Acrobat Reader or other PDF viewer.
EPrint Type:Conference or Workshop Item (Oral)
Project Keyword:Project Keyword UNSPECIFIED
Subjects:Machine Vision
Learning/Statistics & Optimisation
ID Code:1982
Deposited By:Alexei Pozdnoukov
Deposited On:08 January 2006