Custom keypoint detection model through vision api

Hi there,

I have a custom keypoint detection model and want to use it via vision's CoremlRequest API. Here's some complication for input and output:

For input My model expect 512x512 a image. Which would be resized and padded from a 1920x1080 frame. I use the .scaleToFit option, but can I also specify the color used for padding?

For output:

  1. My model output a CoreMLFeatureValueObservation, can I have it output in a format vision recognizes? such as joints/keypoints
  2. If my model is able to output in a format vision recognizes, would it take care to restoring the coordinates back to the original frame? (undo the padding) If not, how do I restore it from .scaletofit option?

Best,

Custom keypoint detection model through vision api
 
 
Q