Are you sure you want to delete this task? Once this task is deleted, it cannot be recovered.
zangxh 6f872d69f8 | 2 years ago | |
---|---|---|
configs | 2 years ago | |
fastreid | 2 years ago | |
projects | 2 years ago | |
tests | 2 years ago | |
LICENSE | 2 years ago | |
README.md | 2 years ago | |
test.py | 2 years ago |
Implementation of the proposed LDS. For the preprint version, please refer to [Arxiv].
Here is a brief instruction for installing the experimental environment.
# install virtual envs
$ conda create -n LDS python=3.7 -y
$ conda activate LDS
# install pytorch 1.6.0 and other dependencies
The pre-trained vit model can be downloaded in this link(code: 7qup) and should be put in the /home/[USER]/.cache/torch/checkpoints/
directory.
$ python projects\LDS\train_net.py --config-file projects/LDS/configs/Market1501/LDS_3Branch_mutual.yml
The results of Market1501, DukeMTMC-reID, MSMT17, P-DukeMTMC-reID, and Occluded DukeMTMC are provided below.
Model | Rank-1@Market1501 | Rank-1@DukeMTMC-reID | Rank-1@MSMT17 | Rank-1@P-DukeMTMC-reID | Rank-1@Occluded DukeMTMC |
---|---|---|---|---|---|
LDS | 95.84 | 91.56 | 86.54 | 91.96 | 64.39 |
You can download these models in this link(code: huv8) and put them in the WEIGHT
directory of the yml file. Then use the command below to evaluate them.
$ python projects\LDS\train_net.py --config-file projects/LDS/configs/Market1501/LDS_3Branch_mutual_test.yml
This repository is built upon the repository fast-reid.
If you find this project useful for your research, please kindly cite:
@article{zang2021learning,
author = {Xianghao Zang and Ge Li and Wei Gao and Xiujun Shu},
title = {Learning to disentangle scenes for person re-identification},
journal = {Image and Vision Computing},
volume = {116},
pages = {104330},
year = {2021},
issn = {0262-8856},
doi = {10.1016/j.imavis.2021.104330}
}
This repository is released under the GPL-2.0 License as found in the LICENSE file.
Learning to Disentangle Scenes for Person Re-identification
C Python Cython
Dear OpenI User
Thank you for your continuous support to the Openl Qizhi Community AI Collaboration Platform. In order to protect your usage rights and ensure network security, we updated the Openl Qizhi Community AI Collaboration Platform Usage Agreement in January 2024. The updated agreement specifies that users are prohibited from using intranet penetration tools. After you click "Agree and continue", you can continue to use our services. Thank you for your cooperation and understanding.
For more agreement content, please refer to the《Openl Qizhi Community AI Collaboration Platform Usage Agreement》