I'm exploring machine learning with Pytorch and Tensorflow. I've created my own models that I'm currently integrating with ReShade, a tool allowing to add inject custom post-processing to video games.
See below for a description of the models I've implemented.
Tensorflow model that addresses aliasing in images. Basically, it is an anti-aliasing neural network that, given an image, produces a similar image but without its aliasing. It has been mainly trained for Final Fantasy XIV.
I've created this shader to try my hand at converting a Tensorflow model to a shader.