Using our method, given a support image and skeleton we can
Locate and predict keypoints on images for pose estimation