Learning Coupled Classifiers with RGB images for RGB-D object recognition

作者：

Highlights：

• We aim to use RGB images to help recognizing RGB-D images.

• The correlation between the two target classifiers is utilized.

• The correlation between source and target domains is utilized.

• A unified framework is presented to learn the classifier parameters.

• Extensive experiment results show the astonishing results.

摘要

Highlights•We aim to use RGB images to help recognizing RGB-D images.•The correlation between the two target classifiers is utilized.•The correlation between source and target domains is utilized.•A unified framework is presented to learn the classifier parameters.•Extensive experiment results show the astonishing results.

论文关键词：Object recognition,Depth images,Domain adaptation,L2,1 norm

论文评审过程：Received 7 April 2016, Revised 13 August 2016, Accepted 18 August 2016, Available online 20 August 2016, Version of Record 29 August 2016.

论文官网地址：https://doi.org/10.1016/j.patcog.2016.08.016