Cross-Dataset Person Re-Identification via Unsupervised Pose Disentanglement and Adaptation

[Paper] Pytorch implementation for our ICCV 2019 paper.

Prerequisites

• Python 3
• Pytorch (We run the code under version 0.4.1)

Getting Started

Installation

• Clone this repo:

git clone https://github.com/yujheli/PDA-Net
cd CPD-GAN/

• Install dependencies. You can install all the dependencies by:

pip install -r requirement.txt

Datasets

We conduct experiments on Market1501, DukeMTMC-reID, CUHK03 datasets. We need pose landmarks for each dataset during training, so we generate the pose files by Realtime Multi-Person Pose Estimation. And the raw datasets have been preprocessed by the code in open-reid.

Acknowledgements

Our code is HEAVILY borrowed and modified from FD-GAN and open-reid.

Citation

Please cite our paper if you find the code useful for your research.

@InProceedings{Li_2019_ICCV,
author = {Li, Yu-Jhe and Lin, Ci-Siang and Lin, Yan-Bo and Wang, Yu-Chiang Frank},
title = {Cross-Dataset Person Re-Identification via Unsupervised Pose Disentanglement and Adaptation},
booktitle = {The IEEE International Conference on Computer Vision (ICCV)},
month = {October},
year = {2019}
}