{"id":13444001,"url":"https://github.com/V2AI/Det3D","last_synced_at":"2025-03-20T17:32:42.182Z","repository":{"id":41403756,"uuid":"203170986","full_name":"V2AI/Det3D","owner":"V2AI","description":"World's first general purpose 3D object detection codebse.","archived":false,"fork":false,"pushed_at":"2023-12-19T10:21:31.000Z","size":3670,"stargazers_count":1520,"open_issues_count":21,"forks_count":298,"subscribers_count":41,"default_branch":"master","last_synced_at":"2025-03-09T03:56:40.350Z","etag":null,"topics":["3d-object-detection","autonomous-driving","deep-learning","kitti","nuscenes","object-detection","point-cloud","pytorch"],"latest_commit_sha":null,"homepage":"https://arxiv.org/abs/1908.09492","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/V2AI.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,"governance":null,"roadmap":null,"authors":null}},"created_at":"2019-08-19T12:43:56.000Z","updated_at":"2025-03-06T13:58:31.000Z","dependencies_parsed_at":"2023-12-19T12:07:23.692Z","dependency_job_id":"2acbd0ff-a80a-4bbc-b481-49c12e57498a","html_url":"https://github.com/V2AI/Det3D","commit_stats":null,"previous_names":["poodarchu/det3d"],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/V2AI%2FDet3D","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/V2AI%2FDet3D/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/V2AI%2FDet3D/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/V2AI%2FDet3D/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/V2AI","download_url":"https://codeload.github.com/V2AI/Det3D/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":244169197,"owners_count":20409679,"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","autonomous-driving","deep-learning","kitti","nuscenes","object-detection","point-cloud","pytorch"],"created_at":"2024-07-31T03:02:16.311Z","updated_at":"2025-03-20T17:32:42.149Z","avatar_url":"https://github.com/V2AI.png","language":"Python","funding_links":[],"categories":["Python"],"sub_categories":[],"readme":"# Det3D\n\nA general 3D Object Detection codebase in PyTorch.\n\n## 1. Introduction\n\nDet3D is the first 3D Object Detection toolbox which provides off the box implementations of many 3D object detection algorithms such as PointPillars, SECOND, PIXOR, etc, as well as state-of-the-art methods on major benchmarks like KITTI(ViP) and nuScenes(CBGS). Key features of Det3D include the following aspects:\n\n* Multi Datasets Support: KITTI, nuScenes, Lyft\n* Point-based and Voxel-based model zoo\n* State-of-the-art performance\n* DDP \u0026 SyncBN\n\n\n## 2. Installation\n\nPlease refer to [INSTALATION.md](INSTALLATION.md).\n\n## 3. Quick Start\n\nPlease refer to [GETTING_STARTED.md](GETTING_STARTED.md).\n\n## 4. Model Zoo\n\n### 4.1 nuScenes\n\n|             | mAP  | mATE | mASE | mAOE | mAVE | mAAE | NDS | ckpt |\n| ----------- | ---- | ---- | ---- | ---- | ---- | ---- | ---- | ---- |\n| [CBGS](https://github.com/poodarchu/Det3D/blob/master/examples/cbgs/configs/nusc_all_vfev3_spmiddleresnetfhd_rpn2_mghead_syncbn.py) | 49.9 | 0.335 | 0.256 | 0.323 | 0.251 | 0.197 | 61.3 | [link](https://drive.google.com/drive/folders/1rhamAqegE9iOp18tzQVam4rOMhHjjnRM?usp=sharing) |\n| [PointPillar](examples/point_pillars/configs/nusc_all_point_pillars_mghead_syncbn.py) | 41.8 | 0.363 | 0.264 | 0.377 | 0.288 | 0.198 | 56.0 | [link](https://drive.google.com/drive/folders/1U0bkEQAhcxhDUD42nTCGC0uU0qaTO_Uv?usp=sharing) |\n\nThe original model and prediction files are available in the [CBGS README](https://github.com/poodarchu/Det3D/tree/master/examples/cbgs).\n\n### 4.2 KITTI\n\n### [Second](examples/second/configs/kitti_car_vfev3_spmiddlefhd_rpn1_mghead_syncbn.py) on KITTI(val) Dataset\n\n```\ncar  AP @0.70, 0.70,  0.70:\nbbox AP:90.54, 89.35, 88.43\nbev  AP:89.89, 87.75, 86.81\n3d   AP:87.96, 78.28, 76.99\naos  AP:90.34, 88.81, 87.66\n```\n\n### [PointPillars](examples/point_pillars/configs/kitti_point_pillars_mghead_syncbn.py) on KITTI(val) Dataset\n\n```\t\ncar  AP@0.70,  0.70,  0.70:\nbbox AP:90.63, 88.86, 87.35\nbev  AP:89.75, 86.15, 83.00\n3d   AP:85.75, 75.68, 68.93\naos  AP:90.48, 88.36, 86.58\n```\n\n\n### 4.3 Lyft\n\n* [Lyft Config](https://github.com/poodarchu/Det3D/blob/master/examples/cbgs/configs/lyft_all_vfev3_spmiddleresnetfhd_rpn2_mghead_syncbn.py)\n\n### 4.4 Waymo\n\n\n\n## 5. Functionality\n\n* Models\n  - [x] VoxelNet\n  - [x] SECOND\n  - [x] PointPillars\n* Features\n    - [x] Multi task learning \u0026 Multi-task Learning\n    - [x] Distributed Training and Validation\n    - [x] SyncBN\n    - [x] Flexible anchor dimensions\n    - [x] TensorboardX\n    - [x] Checkpointer \u0026 Breakpoint continue\n    - [x] Self-contained visualization\n    - [x] Finetune\n    - [x] Multiscale Training \u0026 Validation\n    - [x] Rotated RoI Align\n\n\n## 6. TODO List\n* To Be Released\n\n  * [ ] [CGBS](examples/cbgs/configs/lyft_all_vfev3_spmiddleresnetfhd_rpn2_mghead_syncbn.py) on Lyft(val) Dataset\n\n* Models\n  \n  - [ ] PointRCNN\n  - [ ] PIXOR\n\n## 7. Call for contribution.\n* Support Waymo Dataset.\n* Add other 3D detection / segmentation models, such as VoteNet, STD, etc.\n\n## 8. Developers\n\n[Benjin Zhu](https://github.com/poodarchu/) , [Bingqi Ma](https://github.com/a157801)\n\n## 9. License\n\nDet3D is released under the [Apache licenes](LICENES).\n\n## 10. Citation\nDet3D is a derivative codebase of [CBGS](https://arxiv.org/abs/1908.09492), if you find this work useful in your research, please consider cite:\n```\n@article{zhu2019class,\n  title={Class-balanced Grouping and Sampling for Point Cloud 3D Object Detection},\n  author={Zhu, Benjin and Jiang, Zhengkai and Zhou, Xiangxin and Li, Zeming and Yu, Gang},\n  journal={arXiv preprint arXiv:1908.09492},\n  year={2019}\n}\n```\n\n## 11. Acknowledgement\n\n* [mmdetection](https://github.com/open-mmlab/mmdetection) \n* [mmcv](https://github.com/open-mmlab/mmcv)\n* [second.pytorch](https://github.com/traveller59/second.pytorch)\n* [maskrcnn_benchmark](https://github.com/facebookresearch/maskrcnn-benchmark)\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2FV2AI%2FDet3D","html_url":"https://awesome.ecosyste.ms/projects/github.com%2FV2AI%2FDet3D","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2FV2AI%2FDet3D/lists"}