{"id":13440204,"url":"https://github.com/Zengyi-Qin/Weakly-Supervised-3D-Object-Detection","last_synced_at":"2025-03-20T09:32:30.718Z","repository":{"id":37636044,"uuid":"283067232","full_name":"Zengyi-Qin/Weakly-Supervised-3D-Object-Detection","owner":"Zengyi-Qin","description":"Weakly Supervised 3D Object Detection from Point Clouds (VS3D), ACM MM 2020","archived":false,"fork":false,"pushed_at":"2023-03-24T22:43:14.000Z","size":1177,"stargazers_count":102,"open_issues_count":8,"forks_count":18,"subscribers_count":7,"default_branch":"master","last_synced_at":"2024-08-01T03:31:13.127Z","etag":null,"topics":["3d-object-detection","acm-mm-2020","cross-modal","kitti","lidar","monocular","object-proposals","point-cloud","stereo","tensorflow","transfer-learning","unsupervised-learning","unsupervised-object-detection","vs3d","weakly-supervised-detection","ws3d"],"latest_commit_sha":null,"homepage":"","language":"Jupyter Notebook","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/Zengyi-Qin.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null}},"created_at":"2020-07-28T01:48:23.000Z","updated_at":"2024-06-28T16:25:09.000Z","dependencies_parsed_at":"2023-01-21T12:01:38.292Z","dependency_job_id":null,"html_url":"https://github.com/Zengyi-Qin/Weakly-Supervised-3D-Object-Detection","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/Zengyi-Qin%2FWeakly-Supervised-3D-Object-Detection","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Zengyi-Qin%2FWeakly-Supervised-3D-Object-Detection/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Zengyi-Qin%2FWeakly-Supervised-3D-Object-Detection/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Zengyi-Qin%2FWeakly-Supervised-3D-Object-Detection/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/Zengyi-Qin","download_url":"https://codeload.github.com/Zengyi-Qin/Weakly-Supervised-3D-Object-Detection/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":221745274,"owners_count":16873743,"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":["3d-object-detection","acm-mm-2020","cross-modal","kitti","lidar","monocular","object-proposals","point-cloud","stereo","tensorflow","transfer-learning","unsupervised-learning","unsupervised-object-detection","vs3d","weakly-supervised-detection","ws3d"],"created_at":"2024-07-31T03:01:20.659Z","updated_at":"2025-03-20T09:32:30.712Z","avatar_url":"https://github.com/Zengyi-Qin.png","language":"Jupyter Notebook","funding_links":[],"categories":["Jupyter Notebook"],"sub_categories":[],"readme":"## Weakly Supervised 3D Object Detection from Point Clouds (VS3D)\n\n![](demo.png)\n![](demo.gif)\n\n#### Created by [Zengyi Qin](http://www.qinzy.tech/), Jinglu Wang and Yan Lu. The repository contains an implementation of this [ACM MM 2020 Paper](https://arxiv.org/abs/2007.13970). Readers are strongly recommended to create and enter a [virtual environment](https://docs.conda.io/projects/conda/en/latest/user-guide/tasks/manage-environments.html) with Python 3.6 before running the code.\n\n### Quick Demo with Jupyter Notebook\nClone this repository:\n```bash\ngit clone https://github.com/Zengyi-Qin/Weakly-Supervised-3D-Object-Detection.git\n```\nEnter the main folder and run installation:\n```bash\npip install -r requirements.txt\n```\nDownload the [demo data](https://drive.google.com/file/d/1s4G3avlud7H4oqEBpi0GMnL20HjPJ9Wd/view?usp=sharing) to the main folder and run `unzip vs3d_demo.zip`. Readers can try out the quick demo with Jupyter Notebook:\n```bash\ncd core\njupyter notebook demo.ipynb\n```\n\n### Training\nDownload the [Kitti Object Detection Dataset](http://www.cvlibs.net/datasets/kitti/eval_object.php?obj_benchmark=3d) ([image](http://www.cvlibs.net/download.php?file=data_object_image_2.zip), [calib](http://www.cvlibs.net/download.php?file=data_object_calib.zip) and [label](http://www.cvlibs.net/download.php?file=data_object_label_2.zip)) and place them into `data/kitti`. Download the ground planes and front-view XYZ maps from [here](https://drive.google.com/file/d/1SGgIFLK8t7IVf6Cb_yk83NGWzJcx5WwK/view?usp=sharing) and run `unzip vs3d_train.zip`. Download the pretrained teacher network from [here](https://drive.google.com/file/d/1rnTjOg7R8Ino6778OPTxJSvHBzuc50Yv/view?usp=sharing) and run `unzip vs3d_pretrained.zip`. The data folder should be in the following structure:\n```\n├── data\n│   ├── demo\n│   └── kitti\n│       └── training\n│           ├── calib\n│           ├── image_2\n│           ├── label_2\n│           ├── sphere\n│           ├── planes\n│           └── velodyne\n│       ├── train.txt\n│       └── val.txt\n│   └── pretrained\n│       ├── student\n│       └── teacher\n```\nThe `sphere` folder contains the front-view XYZ maps converted from `velodyne` point clouds using the script in `./preprocess/sphere_map.py`. After data preparation, readers can train VS3D from scratch by running:\n```bash\ncd core\npython main.py --mode train --gpu GPU_ID\n```\nThe models are saved in `./core/runs/weights` during training. Reader can refer to `./core/main.py` for other options in training.\n\n### Inference\nReaders can run the inference on KITTI validation set by running:\n```bash\ncd core\npython main.py --mode evaluate --gpu GPU_ID --student_model SAVED_MODEL\n```\nReaders can also directly use the pretrained model for inference by passing `--student_model ../data/pretrained/student/model_lidar_158000`. Predicted 3D bounding boxes are saved in `./output/bbox` in KITTI format. \n\n### Citation\n```\n@article{qin2020vs3d, \n  title={Weakly Supervised 3D Object Detection from Point Clouds}, \n  author={Zengyi Qin and Jinglu Wang and Yan Lu},\n  journal={ACM Multimedia},\n  year={2020}\n}\n```\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2FZengyi-Qin%2FWeakly-Supervised-3D-Object-Detection","html_url":"https://awesome.ecosyste.ms/projects/github.com%2FZengyi-Qin%2FWeakly-Supervised-3D-Object-Detection","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2FZengyi-Qin%2FWeakly-Supervised-3D-Object-Detection/lists"}