Learning context-sensitive similarity by shortest path propagation

• Published in: Pattern recognition Vol. 44; no. 10; pp. 2367 - 2374
• Main Authors: Wang, Jingyan, Li, Yongping, Bai, Xiang, Zhang, Ying, Wang, Chao, Tang, Ning
• Format: Journal Article
• Language: English
• Published: Kidlington Elsevier Ltd 01.10.2011; Elsevier
• Subjects: Algorithms; Applied sciences; Contextual similarity learning; Detection, estimation, filtering, equalization, prediction; Exact sciences and technology; Graph transduction; Information, signal and communications theory; Learning; Manifolds; Pattern recognition; Retrieval; Shape retrieval; Shortest path propagation; Shortest-path problems; Signal and communications theory; Signal processing; Signal, noise; Similarity; Telecommunications and information theory; Graph transduction; Shape retrieval; Shortest path propagation; Contextual similarity learning; Query; Shortest path; Pattern recognition; Context aware; Object recognition; Algorithm; Learning; Object detection; Database; Distance measurement; Target detection
• ISSN: 0031-3203; 1873-5142
• DOI: 10.1016/j.patcog.2011.02.007

In this paper, we introduce a novel shape/object retrieval algorithm shortest path propagation (SSP). Given a query object q and a target database object p, we explicitly find the shortest path between them in the distance manifold of the database objects. Then a new distance measure between q and p is learned based on the database objects on the shortest path to replace the original distance measure. The promising results on both MEPG-7 shape dataset and a protein dataset demonstrate that our method can significantly improve the ranking of the object retrieval. ► GP assumes the similarity between query and target is affected by all other objects. ► We argue it is only affected by a few contextual reference objects. ► These reference objects are the nodes in the shortest path between them. ► Our SSP propagates similarity from query to target along the shortest path.