https://github.com/iCVTEAM/HRCN

Heterogeneous Relational Complement for Vehicle Re-identification, ICCV 2021
https://github.com/iCVTEAM/HRCN

re-identification vehicle

Last synced: 2 months ago
JSON representation

Heterogeneous Relational Complement for Vehicle Re-identification, ICCV 2021

• Host: GitHub
• URL: https://github.com/iCVTEAM/HRCN
• Owner: iCVTEAM
• Created: 2021-09-07T12:13:41.000Z (over 3 years ago)
• Default Branch: master
• Last Pushed: 2021-10-10T05:44:19.000Z (over 3 years ago)
• Last Synced: 2024-08-01T22:37:48.212Z (5 months ago)
• Topics: re-identification, vehicle
• Language: Python
• Homepage:
• Size: 268 KB
• Stars: 20
• Watchers: 5
• Forks: 5
• Open Issues: 2
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

• Vehicle_reID-Collection - code

README

        
# 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](https://drive.google.com/drive/folders/1gDz761-gTF3nLnwU24kDIVzDbCyBJu80?usp=sharing) 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](https://github.com/JDAI-CV/fast-reid).