{"id":18277114,"url":"https://github.com/js2hou/tdcup2022","last_synced_at":"2025-04-05T04:30:26.266Z","repository":{"id":45820344,"uuid":"474238721","full_name":"Js2Hou/TDCUP2022","owner":"Js2Hou","description":"2022年数据挖掘泰迪杯比赛A题代码","archived":false,"fork":false,"pushed_at":"2022-05-01T12:16:52.000Z","size":2870,"stargazers_count":5,"open_issues_count":0,"forks_count":1,"subscribers_count":1,"default_branch":"main","last_synced_at":"2025-03-20T21:40:58.003Z","etag":null,"topics":["tdcup"],"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/Js2Hou.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}},"created_at":"2022-03-26T04:14:51.000Z","updated_at":"2023-08-31T05:59:42.000Z","dependencies_parsed_at":"2022-09-01T04:40:44.338Z","dependency_job_id":null,"html_url":"https://github.com/Js2Hou/TDCUP2022","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/Js2Hou%2FTDCUP2022","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Js2Hou%2FTDCUP2022/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Js2Hou%2FTDCUP2022/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Js2Hou%2FTDCUP2022/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/Js2Hou","download_url":"https://codeload.github.com/Js2Hou/TDCUP2022/tar.gz/refs/heads/main","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":247289382,"owners_count":20914463,"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":["tdcup"],"created_at":"2024-11-05T12:18:08.972Z","updated_at":"2025-04-05T04:30:25.747Z","avatar_url":"https://github.com/Js2Hou.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# TDCUP2022\n\n2022年泰迪杯数据挖掘比赛A题代码，大赛官网 [link](https://www.tipdm.org:10010/#/competition/1481159137780998144/introduce).\n\n赛题任务为农田害虫检测识别，本质为计算机视觉领域的目标检测任务。我们结合Mask R-CNN^1^框架和最新Vision Transformer模型MPViT^2^，设计出适用于农田害虫检测的模型。\n\n\n**项目关键词**：Detectron2、Mask R-CNN、MPViT、Transformer\n\n## 项目结构\n```\nTDCUP2022/\n|-- datasets/\n|   |-- coco/\n|   |\n|   |-- to_coco.py\n\n|-- mpvit/\n|   |-- __init__.py\n|   |-- backbone.py\n|   |-- config.py\n|   |-- dataset_mapper.py\n|   |-- mpvit.py\n|\n|-- output/\n|-- pretrained/\n|\n|-- summmit/\n|   |-- test1/\n|   |-- test2/\n|   |\n|   |-- crop.py\n|   |-- result2to3.py\n|\n|-- scripts/\n|   |-- train.sh  # entrance for training\n|   |-- evaluate.sh\n|\n|-- train_net.py\n|-- predict1.ipynb\n|-- predict2.ipynb\n|-- requirements.txt\n|-- README\n```\n\n- datasets：存放数据集及数据处理脚本\n    - to_coco.py: 将csv格式的标准转换为coco格式\n\n- mpvit：存放骨干模型\n- output：训练输出路径\n- pretrained：预训练模型存储路径\n- summit\n    - test1：测试集1预测结果路径\n    - test2：测试集2预测结果路径\n    - crop.py: 根据预测bbox分类裁剪图片，用于检查检测是否准确\n    - result2to3.py: 根据result2.csv生成result3.csv\n- predict1.ipynb: 用于测试集1的预测\n- predict2.ipynb: 用于测试集2的预测\n\n## 主要工作\n\n- 数据处理\n    - 大赛数据手动筛选\n    - labelme手动标注数据\n    - IP102数据集（未来得及使用）\n- 模型部分\n    - 二分类器判断图片是否包含目标：\n    模型推理比较耗时，为加快速度，训练一个二分类器判断图像是否包含目标，是则检测；否则跳过。\n    - 训练二分类器判断图片是否包含目标，加快测试集推理速度\n    - 分类器纠正检测结果：有害虫外表极其相似，mask rcnn分类错误。重新训练一个分类模型，纠正mask rcnn对检测目标的识别结果。\n    - 检测器：mask rcnn + mpvit\n\n## 参考文献\n[1] He K, Gkioxari G, Dollár P, et al. Mask r-cnn[C]//Proceedings of the IEEE international conference on computer vision. 2017: 2961-2969.\n\n[2] Lee Y, Kim J, Willette J, et al. MPViT: Multi-Path Vision Transformer for Dense Prediction[J]. arXiv preprint arXiv:2112.11010, 2021.\n\n\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fjs2hou%2Ftdcup2022","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fjs2hou%2Ftdcup2022","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fjs2hou%2Ftdcup2022/lists"}