Hi,
In slide 39, we are talking about 3D reconstruction and we use color consistency as one of the losses used to fit the 3D deformable model to the given set of images. In slide 47, we are talking about texture mapping (slide 46). This is done because previous step does not do texture mapping (it only outputs a 3D model without surface texture information). In this step, we use information from images to compute the texture of the 3D model.
