Efficient clustering on Riemannian manifolds: A kernelised random projection approach
作者：
Highlights：
• We propose a kernelised random projection framework for clustering manifold points.
• We present three projection methods conforming to our proposed framework.
• We contrast our proposal to clustering methods on manifolds in various vision tasks.
• We show the proposal obtain significant speed up whilst maintaining the performance.
• We analyse the parameters contributing to the speed up.
摘要
Highlights•We propose a kernelised random projection framework for clustering manifold points.•We present three projection methods conforming to our proposed framework.•We contrast our proposal to clustering methods on manifolds in various vision tasks.•We show the proposal obtain significant speed up whilst maintaining the performance.•We analyse the parameters contributing to the speed up.
论文关键词：Riemannian manifolds,Random projection,Clustering
论文评审过程：Received 19 December 2014, Revised 2 September 2015, Accepted 9 September 2015, Available online 28 September 2015, Version of Record 27 November 2015.
论文官网地址：https://doi.org/10.1016/j.patcog.2015.09.017
