End-to-End Thorough Body Perception for Person Search
Abstract
In this paper, we propose an improved end-to-end multi-branch person search network to jointly optimize person detection, re-identification, instance segmentation, and keypoint detection. First, we build a better and faster base model to extract non-highly correlated feature expression; Second, a foreground feature enhance module is used to alleviate undesirable background noise in person feature maps; Third, we design an algorithm to learn the part-aligned representation for person search. Extensive experiments with ablation analysis show the effectiveness of our proposed end-to-end multi-task model, and we demonstrate its superiority over the state-of-the-art methods on two benchmark datasets including CUHK-SYSU and PRW.
Cite
Text
Tian et al. "End-to-End Thorough Body Perception for Person Search." AAAI Conference on Artificial Intelligence, 2020. doi:10.1609/AAAI.V34I07.6886Markdown
[Tian et al. "End-to-End Thorough Body Perception for Person Search." AAAI Conference on Artificial Intelligence, 2020.](https://mlanthology.org/aaai/2020/tian2020aaai-end/) doi:10.1609/AAAI.V34I07.6886BibTeX
@inproceedings{tian2020aaai-end,
title = {{End-to-End Thorough Body Perception for Person Search}},
author = {Tian, Kun and Huang, Houjing and Ye, Yun and Li, Shiyu and Lin, Jinbin and Huang, Guan},
booktitle = {AAAI Conference on Artificial Intelligence},
year = {2020},
pages = {12079-12086},
doi = {10.1609/AAAI.V34I07.6886},
url = {https://mlanthology.org/aaai/2020/tian2020aaai-end/}
}