Chenghao Jiang, Renkai Wu, Yinghao Liu, Yue Wang, Qing Chang, Pengchen Liang*, and Yuan Fan*
1. Nanjing Medical University, Nanjing, China
2. Shanghai University, Shanghai, China
3. The Affiliated Stomatological Hospital of Nanjing Medical University, Nanjing, China
4. Jiangsu Province Engineering Research Center of Stomatological Translational Medicine, Nanjing, China
5. University of Shanghai for Science and Technology, Shanghai, China
Examples.of.proposed.dataset.segmentation.tasks.mp4
Examples.of.proposed.dataset.classification.tasks.mp4
0. Main Environments.
- python 3.8
- pytorch 1.12.0
1. The proposed datasets (Autooral dataset).
(1) To obtain the Autooral dataset, you need to provide your name, affiliation, reason for applying (one sentence description of your work), and an assurance that you will not share the dataset privately. Specifically, complete the information in the following format and send it to 'wurk@shu.edu.cn' with the subject name 'Autooral dataset request'. We will usually review your request and provide you with a link within 3 days. If you do not register your information as required, your application may fail. Please understand!
Name:
Affiliation:
Reason for applying (one sentence description of your work):
I (the applicant) guarantee that the data will be used only for academic communication and not for any commercial purposes. I (the applicant) guarantee that I will not privately disseminate the data to any public place without the consent of the author.
(2) After getting the Autooral dataset, execute 'Prepare_Autooral.py' for preprocessing to generate the npy file. We also provide annotations for categorization to provide more richness to the study.
2. Train the HF-UNet.
Modify the dataset address in the config_setting.py file to the address where the npy is stored after preprocessing. Then, perform the following operation:
python train.py
- After trianing, you could obtain the outputs in './results/'
3. Test the HF-UNet.
First, in the test.py file, you should change the address of the checkpoint in 'resume_model' and fill in the location of the test data in 'data_path'.
python test.py
- After testing, you could obtain the outputs in './results/'
If you find this repository helpful, please consider citing:
@article{jiang2024high,
title={A high-order focus interaction model and oral ulcer dataset for oral ulcer segmentation},
author={Jiang, Chenghao and Wu, Renkai and Liu, Yinghao and Wang, Yue and Chang, Qing and Liang, Pengchen and Fan, Yuan},
journal={Scientific Reports},
volume={14},
number={1},
pages={20085},
year={2024},
publisher={Nature Publishing Group UK London}
}