Learning non-metric visual similarity for image retrieval

作者：

Highlights：

• To introduce a neural network to model visual similarities

• To propose a regression loss for improving standard metrics in image retrieval

• To improve image retrieval with a simple and effective method

• To design a real end-to-end system for content-based image retrieval

• To show the efficacy of the method via multiple experiments

摘要

•To introduce a neural network to model visual similarities•To propose a regression loss for improving standard metrics in image retrieval•To improve image retrieval with a simple and effective method•To design a real end-to-end system for content-based image retrieval•To show the efficacy of the method via multiple experiments

论文关键词：Image retrieval,Visual similarity,Non-metric learning

论文评审过程：Received 16 April 2018, Revised 12 January 2019, Accepted 19 January 2019, Available online 10 February 2019, Version of Record 20 February 2019.

论文官网地址：https://doi.org/10.1016/j.imavis.2019.01.001