TY - GEN
T1 - A self-immunizing manifold ranking for image retrieval
AU - Wu, Jun
AU - Li, Yidong
AU - Feng, Songhe
AU - Shen, Hong
PY - 2013
Y1 - 2013
N2 - Manifold ranking (MR), as a powerful semi-supervised learning algorithm, plays an important role to deal with the relevance feedback problem in content-based image retrieval (CBIR). However, conventional MR has two main drawbacks: 1) in many cases, it is prone to exploit "unreliable" unlabeled images when deployed in CBIR due to the semantic gap; 2) the performance of MR is quite sensitive to the scale parameter used for calculating the Laplacian matrix. In this work, a self-immunizing MR approach is presented to address the drawbacks. Concretely, we first propose an elastic kNN graph as well as its constructing algorithm to exploit unlabeled images "safely", and then develop a local scaling solution to calculate the Laplacian matrix adaptively. Extensive experiments on 10,000 Corel images show that the proposed algorithm is more effective than the state-of-the-art approaches.
AB - Manifold ranking (MR), as a powerful semi-supervised learning algorithm, plays an important role to deal with the relevance feedback problem in content-based image retrieval (CBIR). However, conventional MR has two main drawbacks: 1) in many cases, it is prone to exploit "unreliable" unlabeled images when deployed in CBIR due to the semantic gap; 2) the performance of MR is quite sensitive to the scale parameter used for calculating the Laplacian matrix. In this work, a self-immunizing MR approach is presented to address the drawbacks. Concretely, we first propose an elastic kNN graph as well as its constructing algorithm to exploit unlabeled images "safely", and then develop a local scaling solution to calculate the Laplacian matrix adaptively. Extensive experiments on 10,000 Corel images show that the proposed algorithm is more effective than the state-of-the-art approaches.
KW - Content-based image retrieval
KW - Elastic kNN graph
KW - Local scaling
KW - Relevance feedback
KW - Self-immunizing manifold ranking
UR - http://www.scopus.com/inward/record.url?scp=84893570722&partnerID=8YFLogxK
U2 - 10.1007/978-3-642-37456-2_36
DO - 10.1007/978-3-642-37456-2_36
M3 - Conference contribution
AN - SCOPUS:84893570722
SN - 9783642374555
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 426
EP - 436
BT - Advances in Knowledge Discovery and Data Mining - 17th Pacific-Asia Conference, PAKDD 2013, Proceedings
T2 - 17th Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2013
Y2 - 14 April 2013 through 17 April 2013
ER -