TY - CHAP
T1 - Small sample size performance of evolutionary algorithms for adaptive image retrieval
AU - Stejić, Zoran
AU - Takama, Yasufumi
AU - Hirota, Kaoru
PY - 2004
Y1 - 2004
N2 - We evaluate the small sample size (SSS) performance of evolutionary algorithms (EAs) for relevance feedback (RP) in image retrieval. We focus on the requirement to learn the user's information need based on a small - between 2 and 25 - number of positive and negative training images. Despite this being a fundamental requirement, none of the existing works dealing with EAs for RF systematically evaluates their SSS performance. To address this issue, we compare four variants of EAs for RF. Common for all variants is the hierarchical, region-based image similarity model, with region and feature weights as parameters. The difference between the variants is in the objective function of the EA used to adjust the model parameters. The objective functions include: (O-l) precision; (O-2) average rank; (O-3) ratio of within-class (i.e., positive images) and between-class (i.e., positive and negative images) scatter; and (O-4) combination of O-2 and O-3. We note that unlike O-l and O-2 - O-3 and O-4 are not used in any of the existing works dealing with EAs for RF. The four variants are evaluated on five test databases, containing 61,895 general-purpose images, in 619 semantic categories. Results of the evaluation reveal that variants with objective functions O-3 and O-4 consistently outperform those with O-l and O-2. Furthermore, comparison with the representative of the existing RF methods shows that EAs are both effective and efficient approaches for SSS learning in region-based image retrieval.
AB - We evaluate the small sample size (SSS) performance of evolutionary algorithms (EAs) for relevance feedback (RP) in image retrieval. We focus on the requirement to learn the user's information need based on a small - between 2 and 25 - number of positive and negative training images. Despite this being a fundamental requirement, none of the existing works dealing with EAs for RF systematically evaluates their SSS performance. To address this issue, we compare four variants of EAs for RF. Common for all variants is the hierarchical, region-based image similarity model, with region and feature weights as parameters. The difference between the variants is in the objective function of the EA used to adjust the model parameters. The objective functions include: (O-l) precision; (O-2) average rank; (O-3) ratio of within-class (i.e., positive images) and between-class (i.e., positive and negative images) scatter; and (O-4) combination of O-2 and O-3. We note that unlike O-l and O-2 - O-3 and O-4 are not used in any of the existing works dealing with EAs for RF. The four variants are evaluated on five test databases, containing 61,895 general-purpose images, in 619 semantic categories. Results of the evaluation reveal that variants with objective functions O-3 and O-4 consistently outperform those with O-l and O-2. Furthermore, comparison with the representative of the existing RF methods shows that EAs are both effective and efficient approaches for SSS learning in region-based image retrieval.
UR - http://www.scopus.com/inward/record.url?scp=35048843434&partnerID=8YFLogxK
U2 - 10.1007/978-3-540-27814-6_10
DO - 10.1007/978-3-540-27814-6_10
M3 - Chapter
AN - SCOPUS:35048843434
SN - 3540225390
SN - 9783540225393
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 51
EP - 59
BT - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
A2 - Enser, Peter
A2 - Kompatsiaris, Yiannis
A2 - O’Connor, Noel E.
A2 - Smeaton, Alan F.
A2 - Smeulders, Arnold W. M.
PB - Springer Verlag
ER -