The Bayesian image retrieval system, PicHunter: Theory, implementation, and psychophysical experiments

DMCA

The Bayesian image retrieval system, PicHunter: Theory, implementation, and psychophysical experiments (2000)

Cached

Download Links

Other Repositories/Bibliography

DBLP

by Ingemar J. Cox , Matt L. Miller , Thomas P. Minka , Thomas V. Papathomas , Peter N. Yianilos

Venue:	IEEE TRANSACTIONS ON IMAGE PROCESSING
Citations:	226 - 2 self

BibTeX

@ARTICLE{Cox00thebayesian,
    author = {Ingemar J. Cox and Matt L. Miller and Thomas P. Minka and Thomas V. Papathomas and Peter N. Yianilos},
    title = {The Bayesian image retrieval system, PicHunter: Theory, implementation, and psychophysical experiments},
    journal = {IEEE TRANSACTIONS ON IMAGE PROCESSING},
    year = {2000},
    volume = {9},
    number = {1},
    pages = {20--37}
}

OpenURL

Abstract

This paper presents the theory, design principles, implementation, and performance results of PicHunter, a prototype content-based image retrieval (CBIR) system that has been developed over the past three years. In addition, this document presents the rationale, design, and results of psychophysical experiments that were conducted to address some key issues that arose during PicHunter’s development. The PicHunter project makes four primary contributions to research on content-based image retrieval. First, PicHunter represents a simple instance of a general Bayesian framework we describe for using relevance feedback to direct a search. With an explicit model of what users would do, given what target image they want, PicHunter uses Bayes’s rule to predict what is the target they want, given their actions. This is done via a probability distribution over possible image targets, rather than by refining a query. Second, an entropy-minimizing display algorithm is described that attempts to maximize the information obtained from a user at each iteration of the search. Third, PicHunter makes use of hidden annotation rather than a possibly inaccurate/inconsistent annotation structure that the user must learn and make queries in. Finally, PicHunter introduces two experimental paradigms to quantitatively evaluate the performance of the system, and psychophysical experiments are presented that support the theoretical claims.

Keyphrases

psychophysical experiment bayesian image retrieval system experimental paradigm theoretical claim relevance feedback probability distribution design principle inaccurate inconsistent annotation structure general bayesian framework target image hidden annotation performance result primary contribution content-based image retrieval key issue explicit model simple instance prototype content-based image retrieval entropy-minimizing display algorithm pichunter project bayes rule possible image target

DMCA