What Makes Objects Similar: A Unified Multi-Metric Learning Approach

• Published in: IEEE transactions on pattern analysis and machine intelligence 2019-05, Vol.41 (5), p.1257-1270
• Main Authors: Ye, Han-Jia, Zhan, De-Chuan, Jiang, Yuan, Zhou, Zhi-Hua
• Format: Article
• Language: English
• Subjects: Correlation; Couplings; Distance metric learning; Feature extraction; Indexes; Learning; Linkages; Measurement; multi-metric learning; semantic; Semantics; similarity measures; Symmetric matrices
• ISSN: 0162-8828; 1939-3539; 2160-9292
• DOI: 10.1109/TPAMI.2018.2829192

Linkages are essentially determined by similarity measures that may be derived from multiple perspectives. For example, spatial linkages are usually generated based on localities of heterogeneous data. Semantic linkages, however, can come from even more properties, such as different physical meanings behind social relations. Many existing metric learning models focus on spatial linkages but leave the rich semantic factors unconsidered. We propose a Unified Multi-Metric Learning ( Um ^2 2 l ) framework to exploit multiple types of metrics with respect to overdetermined similarities between linkages. In Um ^2 2 l , types of combination operators are introduced for distance characterization from multiple perspectives, and thus can introduce flexibilities for representing and utilizing both spatial and semantic linkages. Besides, we propose a uniform solver for Um ^2 2 l , and the theoretical analysis reflects the generalization ability of Um ^2 2 l as well. Extensive experiments on diverse applications exhibit the superior classification performance and comprehensibility of Um ^2 2 l . Visualization results also validate its ability to physical meanings discovery.