/3D-Structure-guided-Network-for-Tooth-Alignment-in-2D-Photograph

[BMVC 2023] 3D Structure-guided Network for Tooth Alignment in 2D Photograph

Primary LanguagePythonApache License 2.0Apache-2.0

3D Structure-guided Network for Tooth Alignment in 2D Photograph (BMVC 2023)

This repository includes our code for the paper '3D Structure-guided Network for Tooth Alignment in 2D Photograph' in BMVC 2023.

Method overview

High-Level structure

The code is organized as follows:

How to Use

Get started

We run with Python 3.7 on Windows, you can set up a conda environment with all dependencies according to Code/requirements.txt.

Download model weights

  • Refer to the links in Code/Stage2/ckpt/download_ckpt.txt, download the model weights and put as Code/Stage2/ckpt/ckpt_contour2contour_mixed_v2_ContourSegm_4000.pth.
  • Refer to the links in Code/Stage3/ckpt/download_ckpt.txt, download the model weights and put as Code/Stage3/ckpt/ckpt_contour2tooth_v2_ContourSegm_facecolor_lightcolor_10000.pth.

Prepare data

Prepare some facial photographs for testing and then put them under path Data/. Here Data/case1.jpg is an example.

Usage

Simply use the following command to run our code. You will see the results in Output/prediction and Output/processing.

   cd Code
   python main.py -i ../Data/case1.jpg

Citation

If our code or models help your work, please cite our paper:

@inproceedings{Dou_2023_BMVC,
author    = {Yulong Dou and Lanzhuju Mei and Dinggang Shen and Zhiming Cui},
title     = {3D Structure-guided Network for Tooth Alignment in 2D Photograph},
booktitle = {34th British Machine Vision Conference 2023, {BMVC} 2023, Aberdeen, UK, November 20-24, 2023},
publisher = {BMVA},
year      = {2023},
url       = {https://papers.bmvc2023.org/0322.pdf}
}