Multi-modal uniform deep learning for RGB-D person re-identification

作者：

Highlights：

• To our best knowledge, we first attempt to use deep learning techniques for RGB-D person re-identification.

• We propose a multi-modal uniform deep learning method to extract the anthropometric and appearance features from RGB-D images for person re-identification.

• Experimental results on two RGB-D person re-identification datasets including the Kinect-REID dataset and the RGBD-ID dataset are presented to show the efficiency and robustness of our proposed approach.

摘要

•To our best knowledge, we first attempt to use deep learning techniques for RGB-D person re-identification.•We propose a multi-modal uniform deep learning method to extract the anthropometric and appearance features from RGB-D images for person re-identification.•Experimental results on two RGB-D person re-identification datasets including the Kinect-REID dataset and the RGBD-ID dataset are presented to show the efficiency and robustness of our proposed approach.

论文关键词：Person re-identification,Deep learning,Multi-model learning

论文评审过程：Received 25 January 2017, Revised 10 May 2017, Accepted 30 June 2017, Available online 4 July 2017, Version of Record 17 August 2017.

论文官网地址：https://doi.org/10.1016/j.patcog.2017.06.037