Code of our paper: {"}Interpretable Neural Networks for Video Separation: Deep Unfolding RPCA with Foreground Masking{"}.Contains the ROMAN-S and ROMAN-R models.