Parametric local multiview hamming distance metric learning

作者：

Highlights：

• We first propose a local (asymmetric) multiview hamming distance metric by exploring a broad class of local multimodal hash functions designed to preserve the semantic structure.

• Local hash functions are approximated with theoretical guarantee, which makes them tractable for large-scale datasets.

• Efficient local hamming metric learning algorithm with weak supervision information is support in a principled manner.

摘要

•We first propose a local (asymmetric) multiview hamming distance metric by exploring a broad class of local multimodal hash functions designed to preserve the semantic structure.•Local hash functions are approximated with theoretical guarantee, which makes them tractable for large-scale datasets.•Efficient local hamming metric learning algorithm with weak supervision information is support in a principled manner.

论文关键词：Metric learning,Hamming distance,Hash function learning

论文评审过程：Received 16 November 2016, Revised 27 March 2017, Accepted 8 June 2017, Available online 5 July 2017, Version of Record 21 November 2017.

论文官网地址：https://doi.org/10.1016/j.patcog.2017.06.018