Image classification based on pyramid histogram of topics

Author

Lu, Fuxiang ; Yang, Xiaokang ; Zhang, Rui ; Yu, Songyu

Author_Institution

Shanghai Key Lab. of Digital Media Process. & Transm., Shanghai Jiao Tong Univ., Shanghai, China

fYear

2009

fDate

June 28 2009-July 3 2009

Firstpage

398

Lastpage

401

Abstract

In this paper we propose PHOTO (pyramid histogram of topics), a new representation for image classification. We partition the image into hierarchical cells and learn the topic histogram using pLSA over each cell with EM algorithm. Then we concatenate the topic histograms over the cells at all levels to form a ldquolongrdquo vector, i.e. pyramid histogram of topics. Finally AdaBoost classifiers are used to select the topics most discriminative for class recognition. Experimental results on two diverse databases show that our method performs significantly better than general topic representation.

Keywords

expectation-maximisation algorithm; image classification; learning (artificial intelligence); probability; statistical analysis; AdaBoost classifier; EM; PHOTO; class recognition; expectation maximisation algorithm; image classification; pLSA; probabilistic latent semantic analysis; pyramid histogram-of-topic; support vector machine; topic histogram learning; Histograms; Image classification; Image communication; Image databases; Image representation; Layout; Linear discriminant analysis; Partitioning algorithms; Support vector machine classification; Support vector machines; AdaBoost; Image classification; PHOTO; pLSA;

fLanguage

English

Publisher

ieee

Conference_Titel

Multimedia and Expo, 2009. ICME 2009. IEEE International Conference on

Conference_Location

New York, NY

ISSN

1945-7871

Print_ISBN

978-1-4244-4290-4

Electronic_ISBN

1945-7871

Type

conf

DOI

10.1109/ICME.2009.5202518

Filename

5202518