This repository contains a highly configurable two-stage-tracker that adjusts to different deployment scenarios. It can jointly perform multiple object tracking and instance segmentation (MOTS). The detections generated by YOLOv8, a family of object detection architectures and models pretrained on the COCO dataset, are passed to the tracker of your choice. Supported ones at the moment are: StrongSORThttps://arxiv.org/abs/2202.13514[] OSNethttps://arxiv.org/abs/1905.00953[], OCSORThttps://arxiv.org/abs/2203.14360[] and ByteTrackhttps://arxiv.org/abs/2110.06864[]. They can track any object that your Yolov8 model was trained to detect.
Why using this tracking toolbox?
Everything is designed with simplicity and flexibility in mind. We don’t hyperfocus on results on a single dataset, we prioritize real-world results. If you don’t get good tracking results on your custom dataset with the out-of-the-box tracker configurations, use the evolve.py script for tracker hyperparameter tuning.
Installation
git clone --recurse-submodules https://raw.githubusercontent.com/mikel-brostrom/yolov8_tracking.git
= clone recursively
cd yolov8_tracking
pip install -r requirements.txt
= install dependencies
----+++<details>++<details>++<details>++++++<summary>+++Custom object detection architecture+++</summary>+++ The trackers provided in this repo can be used with other object detectors than Yolov8. Make sure that the output of your detector has the following format: ```bash (x1,y1, x2, y2, obj, cls0, cls1, \..., clsn) ``` pass this directly to the tracker here: https://raw.githubusercontent.com/mikel-brostrom/Yolov5_StrongSORT_OSNet/a4bc0c38c33023fab9e5481861d9520eb81e28bc/track.py#L189+++</details>+++
== Tracking
[,bash]
$ python track.py --yolo-weights yolov8n.pt
= bboxes only
yolov8-seg.pt
= bboxes + segmentation masks
----++++++++++++++MOT compliant results Can be saved to your experiment folder runs/track/_/ by bash python track.py --source ... --save-txt </details> ++++
@mikel-brostrom
+
Introduction
This repository contains a highly configurable two-stage-tracker that adjusts to different deployment scenarios. It can jointly perform multiple object tracking and instance segmentation (MOTS). The detections generated by YOLOv8, a family of object detection architectures and models pretrained on the COCO dataset, are passed to the tracker of your choice. Supported ones at the moment are: StrongSORThttps://arxiv.org/abs/2202.13514[] OSNethttps://arxiv.org/abs/1905.00953[], OCSORThttps://arxiv.org/abs/2203.14360[] and ByteTrackhttps://arxiv.org/abs/2110.06864[]. They can track any object that your Yolov8 model was trained to detect.
Why using this tracking toolbox?
Everything is designed with simplicity and flexibility in mind. We don’t hyperfocus on results on a single dataset, we prioritize real-world results. If you don’t get good tracking results on your custom dataset with the out-of-the-box tracker configurations, use the
evolve.pyscript for tracker hyperparameter tuning.Installation
$ python track.py --yolo-weights yolov8n.pt = bboxes only yolov8-seg.pt = bboxes + segmentation masks ----++++++++++++ ++++ ++
MOT compliant results
Can be saved to your experiment folderruns/track/_/ bybash python track.py --source ... --save-txt