Restoring images of damaged paintings using in-painting. Damaged paintings have discolored patches where the paint has faded or fallen off. This project uses image in-painting to fill and restore these lost regions. The models include context-encoders, GANS, conditional GANS and pixel diffusion.