Dynamic graph fusion label propagation for semi-supervised multi-modality classification

作者：

Highlights：

• The relevance modeling between multi-graph and label describes their interrelation.

• The nonlinear relation of multi-graph can approximate to the data manifold structure.

• Multi-graph fusion integrates into label propagation for classification performance.

摘要

•The relevance modeling between multi-graph and label describes their interrelation.•The nonlinear relation of multi-graph can approximate to the data manifold structure.•Multi-graph fusion integrates into label propagation for classification performance.

论文关键词：Dynamic graph fusion,Label propagation,Semi-supervised learning,Multi-modality classification

论文评审过程：Received 14 November 2016, Revised 12 January 2017, Accepted 7 March 2017, Available online 8 March 2017, Version of Record 10 March 2017.

论文官网地址：https://doi.org/10.1016/j.patcog.2017.03.014