{"id":26701103,"url":"https://github.com/visionxlab/whollywood-jittor","last_synced_at":"2025-03-27T01:13:55.045Z","repository":{"id":277469729,"uuid":"846361449","full_name":"VisionXLab/whollywood-jittor","owner":"VisionXLab","description":"[TPAMI] Wholly Leveraging Diversified-quality Labels for Weakly-supervised Oriented Object Detection","archived":false,"fork":false,"pushed_at":"2025-02-14T03:54:20.000Z","size":4015,"stargazers_count":3,"open_issues_count":0,"forks_count":0,"subscribers_count":1,"default_branch":"main","last_synced_at":"2025-03-27T01:13:49.921Z","etag":null,"topics":[],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"apache-2.0","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/VisionXLab.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE.txt","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":"2024-08-23T03:38:25.000Z","updated_at":"2025-02-17T09:35:28.000Z","dependencies_parsed_at":"2025-02-14T04:40:33.309Z","dependency_job_id":null,"html_url":"https://github.com/VisionXLab/whollywood-jittor","commit_stats":null,"previous_names":["visionxlab/whollywood-jittor"],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/VisionXLab%2Fwhollywood-jittor","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/VisionXLab%2Fwhollywood-jittor/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/VisionXLab%2Fwhollywood-jittor/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/VisionXLab%2Fwhollywood-jittor/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/VisionXLab","download_url":"https://codeload.github.com/VisionXLab/whollywood-jittor/tar.gz/refs/heads/main","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":245761297,"owners_count":20667895,"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":"2025-03-27T01:13:54.369Z","updated_at":"2025-03-27T01:13:55.033Z","avatar_url":"https://github.com/VisionXLab.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Wholly-WOOD\n## Introduction\nWe develop **Wholly-WOOD** (**Wholly** Leveraging Diversified-quality Labels for **W**eakly-supervised **O**riented **O**bject **D**etection), a weakly-supervised OOD framework, capable of wholly leveraging various labeling forms (Points, HBoxes, RBoxes, and their combination) in a unified fashion. By only using HBox for training, our Wholly-WOOD achieves performance very close to that of the RBox-trained counterpart on remote sensing and other areas, which significantly reduces the tedious efforts on labor-intensive annotation for oriented objects.\n\nThis project is the [Jittor](https://github.com/Jittor/jittor) implementation of Wholly-WOOD. The code works with **Jittor 1.3.8.5**. It is modified from [JDet](https://github.com/Jittor/JDet), which is an object detection benchmark mainly focus on oriented object detection. PyTorch version: [Wholly-WOOD (PyTorch)](https://github.com/yuyi1005/whollywood).\n\n## Models\nThis repository contains the Wholly-WOOD model and our series of work on weakly-supervised OOD (i.e. H2RBox, H2RBox-v2, and Point2RBox).\n\n### 1. Wholly-WOOD\nWe can train/test Wholly-WOOD model by:\n```shell\npython tools/run_net.py --config-file=configs/whollywood/whollywood_obb_r50_adamw_fpn_1x_dota.py --task=train\npython tools/run_net.py --config-file=configs/whollywood/whollywood_obb_r50_adamw_fpn_1x_dota.py --task=test\n```\n\n### 2. H2RBox\nWe can train/test H2RBox model by:\n```shell\npython tools/run_net.py --config-file=configs/whollywood/h2rbox_obb_r50_adamw_fpn_1x_dota.py --task=train\npython tools/run_net.py --config-file=configs/whollywood/h2rbox_obb_r50_adamw_fpn_1x_dota.py --task=test\n```\n\n### 3. H2RBox-v2\nWe can train/test H2RBox-v2 model by:\n```shell\npython tools/run_net.py --config-file=configs/whollywood/h2rbox_v2p_obb_r50_adamw_fpn_1x_dota.py --task=train\npython tools/run_net.py --config-file=configs/whollywood/h2rbox_v2p_obb_r50_adamw_fpn_1x_dota.py --task=test\n```\n\n### 4. Point2RBox\nWe can train/test Point2RBox model by:\n```shell\npython tools/run_net.py --config-file=configs/whollywood/point2rbox_obb_r50_adamw_fpn_1x_dota.py --task=train\npython tools/run_net.py --config-file=configs/whollywood/point2rbox_obb_r50_adamw_fpn_1x_dota.py --task=test\n```\n\n## Installation\nRecommended environments:\n\n* System: **Linux** (e.g. Ubuntu/CentOS/Arch), **macOS**, or **Windows Subsystem of Linux (WSL)**\n* Python == 3.10\n* Jittor == 1.3.8.5\n* CPU Compiler: g++ == 11.4.0\n* GPU Library: cuda == 12.3 \u0026 cudnn == 8.9.7.29\n\n**Step 1: Install requirements**\n```shell\ngit clone https://github.com/yuyi1005/whollywood-jittor\ncd whollywood-jittor\npython -m pip install -r requirements.txt\n```\nIf you have any installation problems for Jittor, please refer to [Jittor](https://github.com/Jittor/jittor)\n\n**Step 2: Install Wholly-WOOD**\n \n```shell\ncd whollywood-jittor\n# suggest this \npython setup.py develop\n# or\npython setup.py install\n```\nIf you don't have permission for install, please add ```--user```.\n\n## Datasets\nThe following datasets are supported in JDet, please check the corresponding document before use. \n\nDOTA1.0/DOTA1.5/DOTA2.0 Dataset: [dota.md](docs/dota.md).\n\nFAIR Dataset: [fair.md](docs/fair.md)\n\nSSDD/SSDD+: [ssdd.md](docs/ssdd.md)\n\nYou can also build your own dataset by convert your datas to DOTA format.\n\n## Visualization\nYou can test and visualize results on your own image sets by:\n```shell\npython tools/run_net.py --config-file=configs/whollywood/whollywood_obb_r50_adamw_fpn_1x_dota.py --task=vis_test\n```\nYou can choose the visualization style you prefer, for more details about visualization, please refer to [visualization.md](docs/visualization.md).\n\u003cimg src=\"https://github.com/Jittor/JDet/blob/visualization/docs/images/vis2.jpg?raw=true\" alt=\"Visualization\" width=\"800\"/\u003e\n\n## Citation\n```\n@article{yu2025whollywood,\n  title={Wholly-WOOD: Wholly Leveraging Diversified-quality Labels for Weakly-supervised Oriented Object Detection}, \n  author={Yi Yu and Xue Yang and Yansheng Li and Zhenjun Han and Feipeng Da and Junchi Yan},\n  year={2025},\n  journal={IEEE Transactions on Pattern Analysis and Machine Intelligence},\n}\n```\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fvisionxlab%2Fwhollywood-jittor","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fvisionxlab%2Fwhollywood-jittor","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fvisionxlab%2Fwhollywood-jittor/lists"}