{"id":13443185,"url":"https://github.com/hailanyi/3D-Multi-Object-Tracker","last_synced_at":"2025-03-20T16:30:45.811Z","repository":{"id":38362332,"uuid":"374277599","full_name":"hailanyi/3D-Multi-Object-Tracker","owner":"hailanyi","description":"A project for 3D multi-object tracking","archived":false,"fork":false,"pushed_at":"2023-06-15T05:36:00.000Z","size":25778,"stargazers_count":316,"open_issues_count":12,"forks_count":53,"subscribers_count":5,"default_branch":"master","last_synced_at":"2024-10-28T06:57:42.444Z","etag":null,"topics":["3d-object-tracking","kitti","multi-object-tracking"],"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/hailanyi.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":"2021-06-06T05:40:23.000Z","updated_at":"2024-10-22T10:54:14.000Z","dependencies_parsed_at":"2024-01-18T14:52:23.575Z","dependency_job_id":null,"html_url":"https://github.com/hailanyi/3D-Multi-Object-Tracker","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/hailanyi%2F3D-Multi-Object-Tracker","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/hailanyi%2F3D-Multi-Object-Tracker/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/hailanyi%2F3D-Multi-Object-Tracker/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/hailanyi%2F3D-Multi-Object-Tracker/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/hailanyi","download_url":"https://codeload.github.com/hailanyi/3D-Multi-Object-Tracker/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":244649707,"owners_count":20487471,"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-tracking","kitti","multi-object-tracking"],"created_at":"2024-07-31T03:01:57.225Z","updated_at":"2025-03-20T16:30:43.199Z","avatar_url":"https://github.com/hailanyi.png","language":"Python","funding_links":[],"categories":["Python","Open Source Solutions"],"sub_categories":[],"readme":"\n# 3D Multi-Object Tracker\nThis project is developed for tracking multiple objects in 3D scene. The visualization code is from\n[here](https://github.com/hailanyi/3D-Detection-Tracking-Viewer). This is a re-produced and simplified version of [paper](https://ieeexplore.ieee.org/abstract/document/9352500). We have made some changes to support more detectors and adapt the HOTA metric.\nWe also provided a simple [guide](https://github.com/hailanyi/3D-Multi-Object-Tracker/blob/master/doc/submit.md) to submit results to the KITTI test set. \n\n![](./doc/demo.gif)\n\n## Features\n* Fast: currently, the codes can achieve 700 FPS using only CPU (not include detection and data op), can perform tracking \non all kitti val sequence in several seconds. \n* Support online, near online and global implementation. \nThe overall framework of design is shown below:\n![](./doc/framework.jpg)\n\n## Kitti Results\n**Car/Pedestrian** tracking results on the Kitti tracking val seq [1,6,8,10,12,13,14,15,16,18,19] \nusing second-iou, point-rcnn and pv-rcnn detections. We also followed the **HOTA** metric, and tuned the parameters by first \nconsidering the **HOTA** performance. \n\n|detector|online|near online (latency=2s)|global|\n|:---:|:---:|:---:|:---:|\n|point-rcnn|76.653|77.576|78.853|\n|second-iou|77.29/48.011|78.17/50.362|78.799/51.20|\n|pv-rcnn|78.289/49.662|79.48/50.824|80.075/51.753|\n\n Online: CA-based KF + greedy matching. \n Near online: online + rescoring tracks in a temporal window.\n Global: online + rescoring tracks globally.\n \n * 2022/9/23 Update CasTrack. The detections are obtained by [CasA detector](https://github.com/hailanyi/CasA) trained on KITTI 3D detection trainval set. Currently **rank first** on the KITTI tracking dataset :fire:!\n\n|Detector|online|near online (latency=2s)|global|detections (training \u0026 testing set)|\n|:---:|:---:|:---:|:---:|:---:|\n|[CasA](https://github.com/hailanyi/CasA)|82.59|83.24|83.34| [detections download link(40M)](https://drive.google.com/file/d/1LaousWNTldOV1IhdcGDRM_UGi5BFWDoN/view?usp=sharing)\n \n * Update VirConvTrack (2023/3/7), the detections are obtained by [VirConv-T detector](https://github.com/hailanyi/VirConv) trained on KITTI 3D detection trainval set.\n\n|Detector|online|near online (latency=2s)|global|detections (training \u0026 testing set)|\n|:---:|:---:|:---:|:---:|:---:|\n|[VirConv](https://github.com/hailanyi/VirConv)|85.40|85.98|86.17| [detections download link(8M)](https://drive.google.com/file/d/1i-SM3RWYLzdXTGRnhjTI7vwkk2CHbh7b/view?usp=sharing)\n \n## Waymo Results\n**Vehicle/Pedestrian/Cyclist** online tracking results, evaluated by **MOTA/L2** metrics of Waymo benchmark. \n\n* 2022/9/23  Update [CasTrack-waymo](https://github.com/hailanyi/CasTrack-waymo), currently **ranks third** on Waymo tracking benchmark :fire:! You can download the CasA detections from [here](https://drive.google.com/file/d/1xDTXrKf3azvW0m6e3xMV1PluICRDstPk/view?usp=sharing)\n(Please register with waymo before downloading, as the Waymo open dataset is under strict non-commercial license).\n\n|detector|set|Vehicle_L2|Pedestrian_L2|Cyclist_L2| ALL_NS_L2 |\n|:---:|:---:|:---:|:---:|:---:|:---:|\n|CasA|val set|59.30|62.78|61.72|61.27|\n|CasA|test set|63.66|64.79|59.34|62.60|\n \n## Prepare data \nYou can download the Kitti tracking pose data from [here](https://drive.google.com/drive/folders/1Vw_Mlfy_fJY6u0JiCD-RMb6_m37QAXPQ?usp=sharing), and\nyou can download the point-rcnn, second-iou and pv-rcnn detections from [here](https://drive.google.com/file/d/1zVWFGwRqF_CBP4DFJJa4nBcu-z6kpF1R/view?usp=sharing).\nYou can download the CasA detections(including training \u0026 testing set) from [here](https://drive.google.com/file/d/1LaousWNTldOV1IhdcGDRM_UGi5BFWDoN/view?usp=sharing).\n\nTo run this code, you should organize Kitti tracking dataset as below:\n```\n# Kitti Tracking Dataset       \n└── kitti_tracking\n       ├── testing \n       |      ├──calib\n       |      |    ├──0000.txt\n       |      |    ├──....txt\n       |      |    └──0028.txt\n       |      ├──image_02\n       |      |    ├──0000\n       |      |    ├──....\n       |      |    └──0028\n       |      ├──pose\n       |      |    ├──0000\n       |      |    |    └──pose.txt\n       |      |    ├──....\n       |      |    └──0028\n       |      |         └──pose.txt\n       |      ├──label_02\n       |      |    ├──0000.txt\n       |      |    ├──....txt\n       |      |    └──0028.txt\n       |      └──velodyne\n       |           ├──0000\n       |           ├──....\n       |           └──0028      \n       └── training # the structure is same as testing set\n              ├──calib\n              ├──image_02\n              ├──pose\n              ├──label_02\n              └──velodyne \n```\nDetections\n```\n└── point-rcnn\n       ├── training\n       |      ├──0000\n       |      |    ├──000001.txt\n       |      |    ├──....txt\n       |      |    └──000153.txt\n       |      ├──...\n       |      └──0020\n       └──testing \n```\n\n## Requirements\n```\npython3\nnumpy\nopencv\nyaml\n```\n\n## Quick start\n* Please modify the dataset path and detections path in the [yaml file](./config/online/pvrcnn_mot.yaml) \nto your own path.\n* Then run ``` python3 kitti_3DMOT.py config/online/pvrcnn_mot.yaml``` \n* The results are automatically saved to ```evaluation/results/sha_key/data```, and \nevaluated by HOTA metrics.\n\n## Notes\nThe evaluation codes are copied from [Kitti](https://github.com/JonathonLuiten/TrackEval).\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fhailanyi%2F3D-Multi-Object-Tracker","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fhailanyi%2F3D-Multi-Object-Tracker","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fhailanyi%2F3D-Multi-Object-Tracker/lists"}