标题：SOMH: A Self-Organizing map based topology preserving hashing method
作者：Liang, Xiao-Long ;Xu, Xin-Shun ;Cui, Lizhen ;Guo, Shanqing ;Wang, Xiao-Lin
作者机构：[Liang, Xiao-Long ;Xu, Xin-Shun ;Cui, Lizhen ;Guo, Shanqing ;Wang, Xiao-Lin ] School of Computer Science and Technology, Shandong University, Jinan, C 更多
会议名称：22nd International Conference on MultiMedia Modeling, MMM 2016
会议日期：4 January 2016 through 6 January 2016
来源：Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
关键词：Approximate nearest neighbor search; Hashing; Media research; Self-organizing map
摘要：Hashing based approximate nearest neighbor search techniques have attracted considerable attention in media search community. An essential problem of hashing is to keep the neighborhood relationship while doing hashing map. In this paper, we propose a self-organizing map based hashing method–SOMH, which cannot only keep similarity relationship, but also preserve topology of data. Specifically, in SOMH, selforganizing map is introduced to map data points into hamming space. In this framework, in order to make it work well on short and long binary codes, we propose a relaxed version of SOMH and a product space SOMH, respectively. For the optimization problem of relaxed SOMH, we also present an iterative solution. To test the performance of SOMH, we conduct experiments on two benchmark datasets–SIFT1M and GIST1M. Experimental results show that SOMH can outperform or is comparable to several state-of-the-arts. © Springer International Publishing Switzerland 2016.