Learning visual models of semantic concepts

Author

Naphade, Milind R. ; Smith, John R.

Author_Institution

IBM Thomas J. Watson Res. Center, Hawthorne, NY, USA

Volume

2

fYear

2003

fDate

14-17 Sept. 2003

Abstract

Statistical machine learning provides a computational framework for mapping low level media features to high level semantics concepts. In this paper we expose the challenges that these techniques face. Using support vector machine (SVM) classification we build models for 34 semantic concepts for the TREC 2002 benchmark corpus. We study the effect of number of examples available for training with respect to their impact on detection. We also examine low level feature fusion as well as parameter sensitivity with SVM classifiers.

Keywords

content-based retrieval; feature extraction; image retrieval; learning (artificial intelligence); multimedia systems; support vector machines; SVM classifiers; TREC 2002 benchmark corpus; computational framework; feature fusion; media features mapping; multimedia content retrieval; parameter sensitivity; semantics concept; statistical machine learning; support vector machine; Benchmark testing; Content based retrieval; Data mining; Face detection; Feature extraction; Feedback; Machine learning; Shape; Support vector machine classification; Support vector machines;

fLanguage

English

Publisher

ieee

Conference_Titel

Image Processing, 2003. ICIP 2003. Proceedings. 2003 International Conference on

ISSN

1522-4880

Print_ISBN

0-7803-7750-8

Type

conf

DOI

10.1109/ICIP.2003.1246734

Filename

1246734