mikel-brostrom / boxmot

BoxMOT: pluggable SOTA tracking modules for segmentation, object detection and pose estimation models
GNU Affero General Public License v3.0
6.74k stars 1.72k forks source link

How Can I use yolov8 custom trained #1168

Closed TuanBao0711 closed 1 year ago

TuanBao0711 commented 1 year ago

Search before asking

Question

I want to use my custom train yolov8 model to tracking some object . How can I use this in scripts?

TuanBao0711 commented 1 year ago

I have a custom trained model yolov8 to detect a UAV, name "UAV_v8.pt". How can I use this to tracking UAV in video? Thank you so much ^^

TuanBao0711 commented 1 year ago

I use: python examples/track.py --tracking-method deepocsort --yolo-model UAV_v8 --show --source video/IR.mp4 =>2023-11-01 16:09:35.931 | ERROR | examples.detectors:get_yolo_inferer:34 - Failed to infer inference mode from yolo model name 2023-11-01 16:09:35.931 | ERROR | examples.detectors:get_yolo_inferer:35 - Your model name has to contain either yolox, yolo_nas or yolov8

mikel-brostrom commented 1 year ago

Rename the model, from: "UAV_v8.pt" to "yolov8_UAV.pt". Then you can run:

python examples/track.py --tracking-method deepocsort --yolo-model yolov8_UAV.pt --show --source video/IR.mp4
TuanBao0711 commented 1 year ago

Thanks you so much, it work