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

Inquiry Regarding Landmark Confidence Scores Marked as 2.0 in YOLO Face Dataset #78

Open
YoungjaeDev opened this issue Dec 24, 2023 · 1 comment

Comments

@YoungjaeDev
Copy link

Hello,
I have been using the YOLO face detection dataset and came across an interesting observation regarding the landmark confidence scores.
I noticed that the confidence scores for landmarks are marked as 2.0 in the labeled data.
Typically, I understand that confidence scores are represented between 0 and 1, so I am curious about the rationale or intent behind this particular notation.
Is this a reflection of a specific rule or preprocessing method in the dataset, or is there another reason for this notation?
I would greatly appreciate any clarification on this matter.

@ashish-roopan
Copy link

From what I understand, the format of data is like this. 0,cx,cy,w,h,kx1,ky1,v1,kx2,ky2,v2.. .Here 2 is used to represent the visibility , if its is 0 then it means the keypoint is not visible. 2 means visible. While calculating the loss, only those keypoints with visiblility 2 is taken

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