HRCN

This repository contains PyTorch codes for the ICCV2021 paper "Heterogeneous Relational Complement for Vehicle Re-identification"

Installation

Requirements

• Linux with python 3.6
• pytorch 1.4.0
• torchvision 0.5.0
• cudatoolkit 10.0

Set up with Conda

cd HRCN
conda env create -f hrcn.yml
conda activate hrcn
pip install -r requirements.txt

Training and Evaluating

Replace the [source_link] with the dataset directory in dataset_soft_link.sh.

Download trained models into the directory model_weight.

cd HRCN
sh dataset_soft_link.sh

# Train in VehicleID, VeRi or VERIWild
sh trainVehicleID.sh
sh trainVeRi.sh
sh trainVERIWild.sh

# Evaluate in VehicleID, VeRi or VERIWild
sh testVehicleID.sh
sh testVeRi.sh
sh testVERIWild.sh

Citation

@InProceedings{Zhao_2021_ICCV,
    author    = {Zhao, Jiajian and Zhao, Yifan and Li, Jia and Yan, Ke and Tian, Yonghong},
    title     = {Heterogeneous Relational Complement for Vehicle Re-Identification},
    booktitle = {Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV)},
    month     = {October},
    year      = {2021},
    pages     = {205-214}
}

Acknowledgment

This repository is based on the implementation of fast-reid.