search

howyoungchen / label-studio-Yolo-backend

This project, named Label-Studio-Yolo-Backend, is an open-source project aimed at providing an Yolov5 assisted labeling backend for Label Studio using label-studio-ml.
MIT License
2 stars 0 forks source link

Need further guidelines #1

Open XM-8JD2 opened 1 year ago

XM-8JD2 commented 1 year ago

"Clone this repository and Copy the python file to the appropriant dia

It can be set up a directory with any name in "Label-Studio-Ml-Backend-Master \ Label_Studio_ml", and then put "label_studio_yolov5x_backend.py" and "YOLOV5S.pt" in this directory?

"Configure your Label Studio Project to use the label-studio-yolo-backnd." What are the specific details?

Based on "https://mmDetection.readthedocs.io/en/laser_guides/label_studio.html"

you can already use “label-studio-ml start mmdetection\projects\LabelStudio\backend_template --with config_file=mmdetection\configs\rtmdet\rtmdet_m_8xb32-300e_coco.py checkpoint_file=mmdetection\work_dirs\rtmdet_m_8xb32-300e_coco_20220719_112220-229f527c.pth device=cpu --port 8003”

Such a instruction running MMDETECTION machine learning back end, so if you can provide a guide similar to this degree, I can easily perform your tutorial

If you can provide such a case, using Labelstudio for semi -automatic labeling and instant training for YOLOV5 will be easier

XM-8JD2 commented 1 year ago

Interesting coincidence: I use GPT4 to learn about this repository