TY - GEN
T1 - Adaptive multi-class metric content-based image retrieval
AU - Peng, Jing
N1 - Publisher Copyright:
© Springer-Verlag Berlin Heidelberg 2000.
PY - 2000
Y1 - 2000
N2 - Relevance feedback methods for content-based image retrieval have shown promise in a variety of image database applications. These techniques assume two (relevant and irrelevant) class relevance feedback. While simple computationally, two class relevance feedback often becomes inadequate in providing suÆcient information to help rapidly improve retrieval performance. We propose a locally adaptive technique for content-based image retrieval that enables relevance feedback to take on multi-class form. We estimate a exible multi-class metric for computing retrievals based on Chi-squared distance analysis. As a result, local data distributions can be suÆciently exploited, whereby rapid performance improvement can be achieved. The eÆcacy of our method is validated and compared against other competing techniques using a number of real world data sets.
AB - Relevance feedback methods for content-based image retrieval have shown promise in a variety of image database applications. These techniques assume two (relevant and irrelevant) class relevance feedback. While simple computationally, two class relevance feedback often becomes inadequate in providing suÆcient information to help rapidly improve retrieval performance. We propose a locally adaptive technique for content-based image retrieval that enables relevance feedback to take on multi-class form. We estimate a exible multi-class metric for computing retrievals based on Chi-squared distance analysis. As a result, local data distributions can be suÆciently exploited, whereby rapid performance improvement can be achieved. The eÆcacy of our method is validated and compared against other competing techniques using a number of real world data sets.
UR - http://www.scopus.com/inward/record.url?scp=84959050043&partnerID=8YFLogxK
U2 - 10.1007/3-540-40053-2_36
DO - 10.1007/3-540-40053-2_36
M3 - Conference contribution
AN - SCOPUS:84959050043
SN - 3540411771
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 407
EP - 418
BT - Advances in Visual Information Systems - 4th International Conference, VISUAL 2000, Proceedings
A2 - Laurini, Robert
PB - Springer Verlag
T2 - 4th International Conference on Visual Information Systems, VISUAL 2000
Y2 - 2 November 2000 through 4 November 2000
ER -