{"id":13784948,"url":"https://github.com/ChenHongruixuan/I3PE","last_synced_at":"2025-05-11T20:31:40.203Z","repository":{"id":169997743,"uuid":"646103103","full_name":"ChenHongruixuan/I3PE","owner":"ChenHongruixuan","description":"[ISPRS J P\u0026RS 2023] Exchange means change: an unsupervised change detection framework based on intra- and inter-image patch exchange","archived":false,"fork":false,"pushed_at":"2024-01-05T07:43:04.000Z","size":4561,"stargazers_count":34,"open_issues_count":1,"forks_count":0,"subscribers_count":3,"default_branch":"master","last_synced_at":"2024-11-17T21:39:15.697Z","etag":null,"topics":["change-detection","convolutional-neural-networks","deep-learning","object-based-image-analysis","python","pytorch","remote-sensing","semi-supervised-learning","single-temporal","unsupervised-learning"],"latest_commit_sha":null,"homepage":"","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/ChenHongruixuan.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}},"created_at":"2023-05-27T09:46:00.000Z","updated_at":"2024-09-09T07:54:08.000Z","dependencies_parsed_at":null,"dependency_job_id":"6812967a-592b-48f6-af6e-d98f1d099c08","html_url":"https://github.com/ChenHongruixuan/I3PE","commit_stats":{"total_commits":8,"total_committers":1,"mean_commits":8.0,"dds":0.0,"last_synced_commit":"3182c2918bd32a5b7dd44dc7ee71fe09ab92ed7b"},"previous_names":["chenhongruixuan/i3pe"],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/ChenHongruixuan%2FI3PE","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/ChenHongruixuan%2FI3PE/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/ChenHongruixuan%2FI3PE/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/ChenHongruixuan%2FI3PE/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/ChenHongruixuan","download_url":"https://codeload.github.com/ChenHongruixuan/I3PE/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":253632025,"owners_count":21939370,"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":["change-detection","convolutional-neural-networks","deep-learning","object-based-image-analysis","python","pytorch","remote-sensing","semi-supervised-learning","single-temporal","unsupervised-learning"],"created_at":"2024-08-03T19:00:54.350Z","updated_at":"2025-05-11T20:31:37.904Z","avatar_url":"https://github.com/ChenHongruixuan.png","language":"Python","funding_links":[],"categories":["Multispectral"],"sub_categories":["Deep Learning"],"readme":"\u003ch1 align=\"center\"\u003eExchange means change: an unsupervised single-temporal change detection framework based on intra- and inter-image patch exchange\u003c/h1\u003e\n\n\u003ch3 align=\"center\"\u003e \u003ca href=\"https://chrx97.com/\"\u003eHongruixuan Chen\u003c/a\u003e, \u003ca href=\"https://github.com/JTRNEO\"\u003eJian Song\u003c/a\u003e, \u003ca href=\"https://scholar.google.com/citations?user=DbTt_CcAAAAJ\u0026hl=zh-CN\"\u003eChen Wu\u003c/a\u003e, \u003ca href=\"https://scholar.google.com/citations?user=Shy1gnMAAAAJ\u0026hl=zh-CN\"\u003eBo Du\u003c/a\u003e, and \u003ca href=\"https://naotoyokoya.com/\"\u003eNaoto Yokoya\u003c/a\u003e\u003c/h3\u003e\n\nThis is an official implementation of I3PE framework in our ISPRS JP\u0026RS 2023 paper: [Exchange means change: an unsupervised single-temporal change detection framework based on intra- and inter-image patch exchange](https://www.sciencedirect.com/science/article/abs/pii/S092427162300309X).\n\u003cdiv align=\"center\"\u003e\n  \u003cimg src=\"./figure/I3PE.PNG\"\u003e\u003cbr\u003e\u003cbr\u003e\n\u003c/div\u003e\n\n## Get started\n### Requirements\nPlease download the following key python packages in advance.\n```\npython==3.6.15\npytorch==1.7.0\nscikit-learn==0.22.1\nscikit-image==0.17.2\nimageio=2.15.0\nnumpy==1.19.5\ntqdm==4.64.1\n```\n\n### Datasets\nTwo large-scale benchmark datasets, \u003ca href=\"https://github.com/liumency/SYSU-CD\"\u003eSYSU dataset\u003c/a\u003e and \u003ca href=\"https://captain-whu.github.io/SCD/\n\"\u003eSECOND dataset\u003c/a\u003e, are used for experiments. Please download them and organize them in the following way. \n\nFor the \u003cstrong\u003eWuhan dataset\u003c/strong\u003e used in our paper, you can also download it here for your own research [\u003ca href=\"https://drive.google.com/file/d/1f9tWouvzwjqf9oujg6BMh-xESzwEefO4/view?usp=drive_link\"\u003eGoogle Drive\u003c/a\u003e], [\u003ca href=\"https://pan.baidu.com/s/1XLPPwfLl1HpSo0kzidIDpQ?pwd=8d27\"\u003eBaidu Cloud\u003c/a\u003e]. \n```\n├── \u003cTHE-ROOT-PATH-OF-DATA\u003e/\n│   ├── SYSU/     \n|   |   ├── train/\n|   |   |   ├── T1/\n|   |   |   ├── T2/\n|   |   |   ├── GT/\n|   |   ├── val/\n|   |   |   |── ...\n|   |   ├── test/\n|   |   |   |── ...\n|   |   \n│   ├── SECOND/     \n|   |   ├── train/\n|   |   |   |── ...\n|   |   ├── test/\n|   |   |   |── ...\n```\n\n### Generate Single-Temporal Training Sets\nTransfer the images in T1 and T2 under original training set to a new folder and rename the images.\n```\npython construct_single_temporal_set.py\n```\n\nGenerate object maps and cluster maps for intra-image patch exchange method in advance.\n```\npython generate_object.py --dataset_path 'your own path here' --obj_num 1000\n```\n```\npython generate_clustering_map.py --dataset_path 'your own path here' --eps 7 --min_samples 10\n```\n\n### Training Change Detectors\nTraining the deep change detector on the single-temporal training sets using both intra- and inter-image patch exchangem methods. \n```\npython train_network_I3PE.py\n```\n\nUnsupervised change detection results of different methods on the test sets:\n\n|               |       SYSU      |                      |       SECOND    |                      |                    \n|:-------------:|:---------------:|:--------------------:|:---------------:|:--------------------:|\n|     Method    |       OA        |          F1          |       OA        |          F1          |\n|      \u003ca href=\"https://github.com/ChenHongruixuan/ChangeDetectionRepository/tree/master/Methodology/Traditional/CVA\"\u003eCVA      |      0.4539     |        0.3492        |      0.4332     |        0.3003        |  \n|     \u003ca href=\"https://github.com/ChenHongruixuan/ChangeDetectionRepository/tree/master/Methodology/Traditional/MAD\"\u003eIRMAD     |      0.6914     |        0.3705        |      0.6829     |        0.3451        |  \n|    \u003ca href=\"https://github.com/ChenHongruixuan/ChangeDetectionRepository/tree/master/Methodology/Traditional/SFA\"\u003e ISFA      |      0.6977     |        0.3695        |      0.7130     |        0.3293        |  \n|     OBCD      |      0.7091     |        0.4046        |      0.7005     |        0.3426        | \n|     \u003ca href=\"https://ieeexplore.ieee.org/document/9669957\"\u003eDCAE      |      0.7636     |        0.4390        |      0.7600     |        0.3340        |  \n|     \u003ca href=\"https://github.com/sudipansaha/dcvaVHROptical\"\u003eDCVA |      0.6995     |        0.4450        |      0.6795     |        0.3681        | \n|     \u003ca href=\"https://github.com/rulixiang/DSFANet\"\u003eDSFA\u003c/a\u003e       |      0.6326     |        0.4125        |      0.5961     |        0.3301        |  \n|  \u003ca href=\"https://github.com/ChenHongruixuan/KPCAMNet\"\u003eKPCA-MNet\u003c/a\u003e  |      0.7084     |        0.4482        |      0.6793     |        0.3670        | \n|     I3PE      |      0.7305     |        0.5547        |      0.7283     |        0.4380        |  \n\n## Citation\nIf this code or dataset contributes to your research, please kindly consider citing our paper :)\n```\n@article{Chen2023Exchange,\n    title = {Exchange means change: An unsupervised single-temporal change detection framework based on intra- and inter-image patch exchange},\n    author = {Hongruixuan Chen and Jian Song and Chen Wu and Bo Du and Naoto Yokoya},\n    journal = {ISPRS Journal of Photogrammetry and Remote Sensing},\n    volume = {206},\n    pages = {87-105},\n    year = {2023},\n    issn = {0924-2716},\n    doi = {https://doi.org/10.1016/j.isprsjprs.2023.11.004}\n}\n```\n\n## Q \u0026 A\n**For any questions, please [contact us.](mailto:Qschrx@gmail.com)**\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2FChenHongruixuan%2FI3PE","html_url":"https://awesome.ecosyste.ms/projects/github.com%2FChenHongruixuan%2FI3PE","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2FChenHongruixuan%2FI3PE/lists"}