Custom raytracer using OptiX deep neural network denoiser

This raytracer was a project I made during my master studies. Features include soft shadowing, GI, reflections, and refraction. The main features that use randomization, being soft shadows and GI, are made using a single sample per pixel. The denoising work is done using the neural network present in OptiX 6.0. The full project can be found on Github and the thesis can be downloaded here