Multi-view manifold learning with locality alignment

作者：

Highlights：

• We learn low-dimensional spaces contain sufficient information of input views.

• We add locality alignment to enhance the discrimination of the latent spaces.

• We propose frameworks under supervised and unsupervised scenarios.

摘要

•We learn low-dimensional spaces contain sufficient information of input views.•We add locality alignment to enhance the discrimination of the latent spaces.•We propose frameworks under supervised and unsupervised scenarios.

论文关键词：Manifold learning,Multi-view learning,Locality alignment

论文评审过程：Received 25 April 2017, Revised 8 January 2018, Accepted 14 January 2018, Available online 31 January 2018, Version of Record 3 February 2018.

论文官网地址：https://doi.org/10.1016/j.patcog.2018.01.012