Hi, can your model generate accurate depth maps estimates for each of the input images, or one depth map view of the scene?