Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Question on creating and processing keypoints for Yolov8 –pose #91

Open
palmcorp opened this issue May 28, 2024 · 7 comments
Open

Question on creating and processing keypoints for Yolov8 –pose #91

palmcorp opened this issue May 28, 2024 · 7 comments

Comments

@palmcorp
Copy link

The process seems to be 1) Run CVAT and define keypoints for image set. 2) run Json2Yolo to convert to a dataset that can be used by yolov8-pose.3) run yolov8-pose. Is this correct?
There are two conversions in JSON2YOLO-master: general_json2yolo.py and labelbox_json2yolo.py. How should these be utilized?
As data export from CVAT of COCO 1.0 keypoints we have :
Annotations
person_keypoints_default.json
Images
Frame0000.PNG

General_json2yolo is looking in ‘annotations’ where it will find person_keypoints_default.json.
Labelbox_jon2yolo looks into "export-2021-06-29T15_25_41.934Z.json"
What should be in ("export-2021-06-29T15_25_41.934Z.json"? Pointing this to the same json as used in general_json2yolo results in new labels, although with just one class.
@pderrenger
Copy link
Member

Hello,

You're on the right track with your process for creating and processing keypoints for YOLOv8-pose. Here's a concise breakdown:

  1. Run CVAT: Define keypoints for your image set.
  2. Convert Annotations: Use general_json2yolo.py to convert your CVAT-exported COCO keypoints JSON to the YOLO format.
  3. Train with YOLOv8-pose: Use the converted dataset to train your model.

Regarding the two scripts in JSON2YOLO-master:

  • general_json2yolo.py: This script is designed to handle general COCO-format JSON files, like your person_keypoints_default.json.
  • labelbox_json2yolo.py: This script is specific to Labelbox JSON exports and expects a different structure, such as "export-2021-06-29T15_25_41.934Z.json".

Since you are using CVAT and have a COCO keypoints JSON, you should use general_json2yolo.py. Ensure your JSON file is correctly formatted and located in the annotations directory.

For further details on dataset formatting and usage, you can refer to our documentation.

If you encounter any issues or need further assistance, feel free to ask! 😊

@palmcorp
Copy link
Author

palmcorp commented May 29, 2024 via email

@pderrenger
Copy link
Member

@palmcorp hi Paul,

Glad to hear that the information on general_json2yolo.py vs. labelbox_json2yolo.py was helpful! If you have any more questions or need further assistance as you proceed, don't hesitate to reach out. Happy training!

@palmcorp
Copy link
Author

palmcorp commented May 29, 2024 via email

@pderrenger
Copy link
Member

Hi Paul,

Switching to YOLOv10 can indeed offer improvements, especially if you're facing configuration challenges with older versions. YOLOv10 brings enhanced features and optimizations that might simplify your setup and potentially yield better results for pose estimation tasks. It's worth considering the upgrade, especially to stay current with the latest advancements and support.

@palmcorp
Copy link
Author

palmcorp commented May 30, 2024 via email

@pderrenger
Copy link
Member

Hi Paul,

Great, feel free to reach out anytime you need assistance. We're here to help!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
2 participants
@palmcorp @pderrenger