LPDGAN

A Dataset and Model for Realistic License Plate Deblurring [Paper Link]

[Haoyan Gong], [Yuzheng Feng], [Zhenrong Zhang], [Xianxu Hou], [Jingxin Liu], [Siqi Huang] and [Hongbin Liu]

Real-World Motion-blurred License Plate Results

Comparison with the other methods.

Citations

BibTeX

Environment

torch>=2.1.1+cu121

Installation

Install Pytorch first. Then,

pip install -r requirements.txt

Download Dataset

• The LPBlur dataset is available at Google Drive and Baidu Netdisk (access code: 7ylj).

How To Train

• Refer to ./main.py for the configuration of the model to train.
• The training command is like

python main.py --mode train --dataroot ./dataset

The training logs and weights will be saved in the ./checkpoints folder.

Acknowledgement

This project is built mainly based on the excellent Swin-Unet and pix2pix codeframe. We appreciate it a lot for their developers.

Contact

If you have any question, please email m.g.haoyan@gmail.com.