{"id":13442610,"url":"https://github.com/LiWentomng/OrientedRepPoints","last_synced_at":"2025-03-20T14:31:40.871Z","repository":{"id":37662321,"uuid":"368755286","full_name":"LiWentomng/OrientedRepPoints","owner":"LiWentomng","description":"The code for “Oriented RepPoints for Aerial Object Detection (CVPR 2022)”","archived":false,"fork":false,"pushed_at":"2023-11-09T05:53:18.000Z","size":69799,"stargazers_count":270,"open_issues_count":40,"forks_count":43,"subscribers_count":5,"default_branch":"main","last_synced_at":"2024-10-28T05:59:26.456Z","etag":null,"topics":["adaptive-points-learning","aerial-image-detection","oriented-object-detection"],"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/LiWentomng.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":"2021-05-19T05:35:39.000Z","updated_at":"2024-10-15T12:42:07.000Z","dependencies_parsed_at":"2024-01-18T14:41:09.358Z","dependency_job_id":"473169fb-7954-4b1e-a7c8-ade2b385acf9","html_url":"https://github.com/LiWentomng/OrientedRepPoints","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/LiWentomng%2FOrientedRepPoints","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/LiWentomng%2FOrientedRepPoints/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/LiWentomng%2FOrientedRepPoints/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/LiWentomng%2FOrientedRepPoints/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/LiWentomng","download_url":"https://codeload.github.com/LiWentomng/OrientedRepPoints/tar.gz/refs/heads/main","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":244630217,"owners_count":20484338,"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":["adaptive-points-learning","aerial-image-detection","oriented-object-detection"],"created_at":"2024-07-31T03:01:47.997Z","updated_at":"2025-03-20T14:31:39.435Z","avatar_url":"https://github.com/LiWentomng.png","language":"Python","funding_links":[],"categories":["Python"],"sub_categories":[],"readme":"# Oriented RepPoints for Aerial Object Detection\n\u003e Wentong Li, Yijie Chen, Kaixuan Hu, Jianke Zhu* ([Arxiv](https://arxiv.org/pdf/2105.11111v4.pdf))\n\n\u003cimg src=\"https://github.com/LiWentomng/OrientedRepPoints/blob/main/docs/overallnetwork.png\" width=\"800px\"\u003e\n\n# \n* Based on OrientedRepPoints detector, the **2nd**  and **3rd** Places are achieved on the Task 2 and Task 1 respectively in the *“2021 challenge of Learning to Understand Aerial Images([LUAI](https://captain-whu.github.io/LUAI2021/tasks.html))”*. **The detailed codes and introductions about it, please refer to this [repository](https://github.com/hukaixuan19970627/OrientedRepPoints_DOTA) and [知乎](https://zhuanlan.zhihu.com/p/422764914)**.\n\n## Update\n* **About the detailed installation, please see this [CSDN Blog](https://blog.csdn.net/SSSlasH/article/details/125255955).** (Thanks for author@SSSlasH of this blog).\n\n* The code for [MMRotate](https://github.com/open-mmlab/mmrotate) is available now.\n \n* [RepPoints](https://github.com/microsoft/RepPoints) + our **APAA** can obtain **+2.5AP** (36.3 to 38.8) improvement with R-50 on **COCO** dataset for general object detection.\n\n\n# Installation\nPlease refer to ![install.md](https://github.com/LiWentomng/OrientedRepPoints/blob/main/docs/install.md) for installation and dataset preparation.\n\n\n# Getting Started \nThis repo is based on ![mmdetection](https://github.com/open-mmlab/mmdetection). Please see ![getting_started.md](https://github.com/LiWentomng/OrientedRepPoints/blob/main/docs/getting_started.md) for the basic usage.\n\n# Results and Models\nThe results on DOTA test set are shown in the table below. More detailed results please see the paper.\n\n  Model| Backbone  |data aug(HSV+Rotation)| mAP | model| log\n ----  | ----- | ------ |------| ------ | ------  \n OrientedReppoints| R-50|  |75.97 |[model](https://drive.google.com/file/d/13c56u9IFRRdHH-YNmQfqb1y11f7xPfCR/view?usp=sharing) | [log](https://drive.google.com/file/d/1_lrj3gV27iM0v95AnSCRHUZDZWkdJFS_/view?usp=sharing)\n OrientedReppoints| R-101| |76.52 |[model](https://drive.google.com/file/d/1otXS3w0LVopsBKxyYbyQhF6mFDtTIJFX/view?usp=sharing) | [log](https://drive.google.com/file/d/1MgJ7A9INaP3iocy1MQSS1SA6gyIvnTJX/view?usp=sharing)\n OrientedReppoints| Swin-Tiny|  √  | 78.11|[model](https://drive.google.com/file/d/1B03dBSXU9GFGRM8XiyQo2aw6yGnCgB8r/view?usp=sharing) | [log](https://drive.google.com/file/d/1lt5UkBPVM7am6asySRWohXSRK_tGwxV8/view?usp=sharing)\n\nNote: \n* The pretrained model--*swin_tiny_patch4_window7_224* of [Swin-Tiny](https://github.com/microsoft/Swin-Transformer) for pytorch1.4.0 is [here](https://drive.google.com/file/d/1ad4lxks68vngs_pCaqs9w_L-fGvtR7nQ/view?usp=sharing).\n* We recommend to use our demo configs with 4 GPUs.\n* The results are performed on the original DOTA images with 1024x1024 patches. \n* The scale jitter is employed during training. More details see the paper.\n\n\nThe mAOE results on DOTA val set are shown in the table below.\n\n  Model| Backbone | mAOE | Download\n ----  | ----- | ------  | ------\n OrientedReppoints| R-50| 5.93° |[model](https://drive.google.com/file/d/1lGHehF57ObkAt0i9FITkp5yS6ULBZQjx/view?usp=sharing)\n\n Note：Orientation error evaluation (mAOE) is calculated on the val subset(only train subset for training).\n\n# Visual results\nThe visualization code for oriented bounding boxes and learning points is ![here](https://github.com/LiWentomng/OrientedRepPoints/blob/main/tools/parse_pkl/show_learning_points_and_boxes.py).\n\n* Oriented bounding box\n\n\u003cimg src=\"https://github.com/LiWentomng/OrientedRepPoints/blob/main/docs/visualization.png\" width=\"800px\"\u003e\n\n\n# Citation\n```shell\n@inproceeding{orientedreppoints,\n\ttitle=\"Oriented RepPoints for Aerial Object Detection.\",\n\tauthor=\"Wentong {Li}, Yijie {Chen}, Kaixuan {Hu}, Jianke {Zhu}.\",\n\tjournal=\"The IEEE Conference on Computer Vision and Pattern Recognition (CVPR)\",\n\tyear=\"2022\"\n}\n```\n\n\n#  Acknowledgements\nHere are some great resources we benefit. We would espeicially thank the authors of:\n\n[MMdetection](https://github.com/open-mmlab/mmdetection)\n\n[RepPoints](https://github.com/microsoft/RepPoints)\n\n[AerialDetection](https://github.com/dingjiansw101/AerialDetection)\n\n[BeyondBoundingBox](https://github.com/sdl-guozonghao/beyondboundingbox)\n\n\n\n\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2FLiWentomng%2FOrientedRepPoints","html_url":"https://awesome.ecosyste.ms/projects/github.com%2FLiWentomng%2FOrientedRepPoints","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2FLiWentomng%2FOrientedRepPoints/lists"}