Are sparse representations really relevant for image classification?

R Rigamonti, Matthew A Brown, V Lepetit

Research output: Chapter in Book/Report/Conference proceedingConference contribution

152 Citations (Scopus)

Abstract

Recent years have seen an increasing interest in sparse representations for image classification and object recog- nition, probably motivated by evidence from the analysis of the primate visual cortex. It is still unclear, however, whether or not sparsity helps classification. In this pa- per we evaluate its impact on the recognition rate using a shallow modular architecture, adopting both standard fil- ter banks and filter banks learned in an unsupervised way. In our experiments on the CIFAR-10 and on the Caltech- 101 datasets, enforcing sparsity constraints actually does not improve recognition performance. This has an impor- tant practical impact in image descriptor design, as enforc- ing these constraints can have a heavy computational cost.
Original languageEnglish
Title of host publication2011 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)
PublisherIEEE
Pages1545-1552
Number of pages8
ISBN (Print)9781457703942
DOIs
Publication statusPublished - Jul 2011
EventIEEE Computer Vision and Pattern Recognition (CVPR) 2011 - Colorado Springs
Duration: 21 Jun 201125 Jun 2011

Conference

ConferenceIEEE Computer Vision and Pattern Recognition (CVPR) 2011
CityColorado Springs
Period21/06/1125/06/11

Fingerprint Dive into the research topics of 'Are sparse representations really relevant for image classification?'. Together they form a unique fingerprint.

Cite this