This project aims to develop a software implementation of an algorithm for generating perspectives from a set of images, so that the new perspectives have different lighting conditions from the original images, as well as producing maps of object surfaces, all achieved by means of neural models of radiance fields.