Open
Description
MediaPipe Solution (you are using)
tasks-vision
Programming language
JavaScript
Are you willing to contribute it
Yes
Describe the feature and the current behaviour/state
The tasks-vision landmark tasks (e..g PoseLandmarker) does not have a selfieMode option (as with @mediapipe/pose
Will this change the current API? How?
Yes, additional option to optionally accept video as selfie mode frames
Who will benefit with this feature?
People combining tasks-vision with other mediapipe tracking that accepts selfie mode.
Please specify the use cases for this feature
We are combining multi-body pose with multi-hand pose data and want to merge and display the same data in the same visualisation. We currently have to treat the two tracking systems differently when drawing to canvas because one accepts selfieMode, one does not!
Any Other info
No response