Building text features for object image classification

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

Abstract

We introduce a text-based image feature and demon- strate that it consistently improves performance on hard object classification problems. The feature is built using an auxiliary dataset of images annotated with tags, down- loaded from the internet. We do not inspect or correct the tags and expect that they are noisy. We obtain the text fea- ture of an unannotated image from the tags of its k-nearest neighbors in this auxiliary collection. A visual classifier presented with an object viewed un- der novel circumstances (say, a new viewing direction) must rely on its visual examples. Our text feature may not change, because the auxiliary dataset likely contains a similar pic- ture. While the tags associated with images are noisy, they are more stable when appearance changes. We test the performance of this feature using PAS- CAL VOC 2006 and 2007 datasets. Our feature performs well, consistently improves the performance of visual ob- ject classifiers, and is particularly effective when the train- ing dataset is small.

Original languageEnglish (US)
Title of host publication2009 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, CVPR Workshops 2009
PublisherIEEE Computer Society
Pages1367-1374
Number of pages8
ISBN (Print)9781424439935
DOIs
StatePublished - Jan 1 2009
Event2009 IEEE Computer Society Conference on Computer Vision and Pattern Recognition - Miami, FL, United States
Duration: Jun 20 2009Jun 25 2009

Publication series

Name2009 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, CVPR Workshops 2009
Volume2009 IEEE Computer Society Conference on Computer Vision and ...

Other

Other2009 IEEE Computer Society Conference on Computer Vision and Pattern Recognition
CountryUnited States
CityMiami, FL
Period6/20/096/25/09

ASJC Scopus subject areas

  • Computer Vision and Pattern Recognition
  • Biomedical Engineering

Fingerprint Dive into the research topics of 'Building text features for object image classification'. Together they form a unique fingerprint.

Cite this