python - Finding keypoints in mediapipe array - Stack Overflow

admin管理员组

文章数量:1122832

I'm currently working with MediaPipe in Python and using it to extract keypoint data from the pose detection model (or any other model such as hand or face detection). However, I'm wondering if it's possible to manually modify or change the keypoint coordinates that are detected by MediaPipe?

Here are some specifics:

I am using the mediapipe.solutions.pose (or other models) and am able to get the landmarks/ keypoints as a list of coordinates (x, y, z values) possibly. For my use case, I would like to modify these coordinates programmatically after they have been detected, potentially for further processing or visualization.

So, my question is:

Can I manually change the keypoint coordinates that are output by MediaPipe? If so, how can I do this in a way that works seamlessly with the MediaPipe pipeline? Is there a method to "set" or "replace" the keypoints instead of just using the ones provided by the detector?

Is it also possible to know how keypoints are mapped in a video or photo (either in 2D or 3D)?

Snippets are provided below for reference:

How do I find the keypoints in array below?

本文标签： pythonFinding keypoints in mediapipe arrayStack Overflow