Improved Statistical Interference Model for Person Re-identification

DOI

10.2991/icmeit-19.2019.12How to use a DOI?

Keywords

Person re-identification, statistical interference, metric learning.

Abstract

Person Re-identification problem is an important and challenging task in computer vision task. Due to the drastic appearance variation caused by misalignment and illumination changing, traditional metric models are failed in similarity measure of pedestrian images. In this paper, a novel metric learning based method is proposed. It establishes a probability inference model based on the probability models of positive pairs and negative pairs. And a balance parameter is proposed in the metric model to deal with the imbalance problem of samples. Finally, experiments are conducted on the VIPeR dataset compared with some metric learning based model. And the test results verified the effectiveness of the proposed model.

Copyright

© 2019, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Linxuan Li
PY  - 2019/04
DA  - 2019/04
TI  - Improved Statistical Interference Model for Person  Re-identification
BT  - Proceedings of the 3rd International Conference on Mechatronics Engineering and Information Technology (ICMEIT 2019)
PB  - Atlantis Press
SP  - 68
EP  - 73
SN  - 2352-538X
UR  - https://doi.org/10.2991/icmeit-19.2019.12
DO  - 10.2991/icmeit-19.2019.12
ID  - Li2019/04
ER  -