Advanced Search

Browse by Discipline

Scientific Societies

E-print Alerts

Add E-prints

E-print Network

  Advanced Search  

Context-Dependent Kernel Design for Object Matching and Recognition Hichem Sahbi

Summary: Context-Dependent Kernel Design for Object Matching and Recognition
Hichem Sahbi
UMR 5141, CNRS
Telecom ParisTech, France
Jean-Yves Audibert, Jaonary Rabarisoa and Renaud Keriven
Certis Lab
ENPC ParisTech, France
The success of kernel methods including support vec-
tor networks (SVMs) strongly depends on the design of ap-
propriate kernels. While initially kernels were designed in
order to handle fixed-length data, their extension to un-
ordered, variable-length data became more than necessary
for real pattern recognition problems such as object recog-
nition and bioinformatics.
We focus in this paper on object recognition using a new
type of kernel referred to as "context-dependent". Objects,
seen as constellations of local features (interest points, re-


Source: Audibert, Jean-Yves - Département d'Informatique, École Normale Supérieure


Collections: Computer Technologies and Information Sciences