SLP real cleaned up and reconstructed images

Cleaned up and calibrated real data addendum: SLP depth images with in-fill for where hair causes noise and where viewframe drops off side of the bed (uncover, cover1, cover2 - 13,620 images), depth renderings for occlusion-free SLP-3Dfits bodies (only_human - 4,545 images), spatial transforms from the camera to the bed, and ground truth reconstructed pressure maps P+ (4,545 maps) and their contact maps (4,545 maps). These are necessary for training BodyPressure models on the SLP real data.