{"id":18276229,"url":"https://github.com/hilab-git/lcovnet-and-kd","last_synced_at":"2025-06-20T22:42:41.793Z","repository":{"id":108580515,"uuid":"566620723","full_name":"HiLab-git/LCOVNet-and-KD","owner":"HiLab-git","description":null,"archived":false,"fork":false,"pushed_at":"2024-08-06T09:00:54.000Z","size":41719,"stargazers_count":27,"open_issues_count":2,"forks_count":2,"subscribers_count":3,"default_branch":"main","last_synced_at":"2025-03-20T21:38:48.445Z","etag":null,"topics":[],"latest_commit_sha":null,"homepage":null,"language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/HiLab-git.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null}},"created_at":"2022-11-16T03:46:17.000Z","updated_at":"2024-12-19T08:21:52.000Z","dependencies_parsed_at":"2024-08-05T10:46:47.508Z","dependency_job_id":"f110c6e9-c631-4dd3-9883-813e02cc8b87","html_url":"https://github.com/HiLab-git/LCOVNet-and-KD","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/HiLab-git%2FLCOVNet-and-KD","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/HiLab-git%2FLCOVNet-and-KD/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/HiLab-git%2FLCOVNet-and-KD/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/HiLab-git%2FLCOVNet-and-KD/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/HiLab-git","download_url":"https://codeload.github.com/HiLab-git/LCOVNet-and-KD/tar.gz/refs/heads/main","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":247284911,"owners_count":20913691,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":[],"created_at":"2024-11-05T12:15:27.528Z","updated_at":"2025-04-05T03:31:17.865Z","avatar_url":"https://github.com/HiLab-git.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# LCOV-Net: A lightweight CNN for 3D image segmentation with knowledge distillation\n[tmi_link]:https://ieeexplore.ieee.org/document/10083150\n[isbi_link]:https://ieeexplore.ieee.org/abstract/document/9434023\n[word_link]:https://www.sciencedirect.com/science/article/abs/pii/S1361841522002705\n[pymic_link]:https://github.com/HiLab-git/PyMIC\n[pymic_example]:https://github.com/HiLab-git/PyMIC_examples\n[baidu_link]:https://pan.baidu.com/s/1HwD1iqHorgXfYXnrChdzIg\n\nThis repository provides the code for the LCOV-Net that was publihsed on [ISBI 2021][isbi_link] and IEEE [TMI 2023][tmi_link]:\n\n* Q. Zhao, L. Zhong, J. Xiao, J. Zhang, Y. Chen, W. Liao, S. Zhang, G. Wang. “Efficient Multi-Organ Segmentation from 3D Abdominal CT Images with Lightweight Network and Knowledge Distillation.” IEEE Transactions on Medical Imaging, 42, no. 9  (2023): 2513-2523.\n\n* Q. Zhao, H. Wang, G. Wang. \"LCOV-NET: A Lightweight Neural Network For COVID-19 Pneumonia Lesion Segmentation From 3D CT Images\", in IEEE ISBI, pp 42-45, 2021.\n\n\n![result](./pic/result.png)\nVisual comparison between different networks for abdominal organ segmentation on the [WORD][word_link] dataset.\n\n![structure](./pic/kd_structure.png)\nOverview of our proposed lightweight LCOV-Net and KD strategies. LCOV-Net is built on our Lightweight Attention-based Convolutional Blocks (LACB-H and LACB-L) to reduce the model size. To improve itsmperformance, we introduce Class-Affinity Knowledge Distillation (CAKD) and Multi-Scale Knowledge Distillation (MSKD) as shown in (c) to effectively distill knowledge from a heavy-weight teacher model to LCOV-Net. Note that for simplicity, the KD losses are only shown for the highest resolution level.\n\n![structure](./pic/lcovnet_structure.png)\nOur proposed LACB for efficient computation.\n\n\n# Dataset\nPlease contact Xiangde (luoxd1996 AT gmail DOT com) for the WORD dataset (**the label of the testing set can be downloaded now [labelTs](https://github.com/HiLab-git/WORD/blob/main/WORD_V0.1.0_labelsTs.zip)**). Two steps are needed to download and access the dataset: **1) using your google email to apply for the download permission ([Goole Driven](https://drive.google.com/drive/folders/16qwlCxH7XtJD9MyPnAbmY4ATxu2mKu67?usp=sharing), [BaiduPan](https://pan.baidu.com/s/1mXUDbUPgKRm_yueXT6E_Kw))**; **2) using your affiliation email to get the unzip password/BaiduPan access code**. We will get back to you within **two days**, **so please don't send them multiple times**. We just handle the **real-name email** and **your email suffix must match your affiliation**. The email should contain the following information:\n\n    Name/Homepage/Google Scholar: (Tell us who you are.)\n    Primary Affiliation: (The name of your institution or university, etc.)\n    Job Title: (E.g., Professor, Associate Professor, Ph.D., etc.)\n    Affiliation Email: (the password will be sent to this email, we just reply to the email which is the end of \"edu\".)\n    How to use: (Only for academic research, not for commercial use or second-development.)\n    \nIn addition, this work is still ongoing, the **WORD** dataset will be extended to larger and more diverse (more patients, more organs, and more modalities, more clinical hospitals' data and MR Images will be considered to include future), any **suggestion**, **comment**, **collaboration**, and **sponsor** are welcome. \n\n# How to use\n1. Install [PyMIC][pymic_link], and add files to Pymic.\n2. Download the pretrained model and example CT images from [Baidu Netdisk][baidu_link] (extract code 9jlj).\n3. Run `./KD/run.sh`. The results will be saved in `./KD/model/kd`.\n\n# How to Cite\nBibTeX entry for this work:\n\n    @article{zhao2023tmi,\n    author={Zhao, Qianfei and Zhong, Lanfeng and Xiao, Jianghong and Zhang, Jingbo and Chen, Yinan and Liao, Wenjun and Zhang, Shaoting and Wang, Guotai},\n    journal={IEEE Transactions on Medical Imaging}, \n    title={Efficient Multi-Organ Segmentation From 3D Abdominal CT Images With Lightweight Network and Knowledge Distillation}, \n    year={2023},\n    volume={42},\n    number={9},\n    pages={2513-2523},\n    doi={10.1109/TMI.2023.3262680}}\n\n    @inproceedings{zhao2021isbi,\n    author={Zhao, Qianfei and Wang, Huan and Wang, Guotai},\n    booktitle={2021 IEEE 18th International Symposium on Biomedical Imaging (ISBI)}, \n    title={LCOV-NET: A Lightweight Neural Network For COVID-19 Pneumonia Lesion Segmentation From 3D CT Images}, \n    year={2021},\n    pages={42-45},\n    doi={10.1109/ISBI48211.2021.9434023}}","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fhilab-git%2Flcovnet-and-kd","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fhilab-git%2Flcovnet-and-kd","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fhilab-git%2Flcovnet-and-kd/lists"}