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.

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

Option to export yolo pose annotations #6791

Open
lovelesh opened this issue Dec 13, 2024 · 2 comments
Open

Option to export yolo pose annotations #6791

lovelesh opened this issue Dec 13, 2024 · 2 comments

Comments

@lovelesh
Copy link

I would like to request for an option to export yolo pose annotation data.
LS already has option to export yolo object detection and segmentation dataset. So it would be really helpful if there is a mechanism to export key point dataset. Currently, I am downloading the JSON file and manually extracting the keypoints using a python script. If the YOLO format is available then it will be easier to export it to other formats for model training.

@makseq
Copy link
Member

makseq commented Dec 30, 2024

There are some opened questions, e.g.

  1. how to group keypoints per person?
  2. how to map keypoint indexes from LS keypoint labels to yolo format?

What logic do you use for your converter?
What do you think about this approach: #6502 ?

@lovelesh
Copy link
Author

Currently, I do not annotate bounding boxes. I just annotate keypoints for my entire dataset.

  • To extract the annotation, I use similar logic for iterating through the LS json file and extracting keypoint values and saving them as a numpy array file. One file per image.
  • My training script takes the image and the numpy array file for training.
  • I do not use YOLO pose as of now. But, I willing to try Export keypoints to yolo format #6502 approach to see if I can generate dataset using single script.

I was hoping if there was a way to add plugin for the LS to generate the entire dataset automatically through web interface.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants