We seek one Postdoctoral Researcher who will explore the introduction of radio component separation specific constraints in deep learning models. The effort will focus on large-scale reconstruction algorithms able to optimize the component separation from SKA interferometry through the coupling of deep-learning-based image priors with blind source separation techniques. The post-doctoral researcher will be located at the premises of FORTH with a strong collaboration with the CosmoStat Laboratory at CEA Saclay. The post-doctoral researcher will be advised by Jean-Luc Starck (FORTH/CEA) and Panagiotis Tsakalides (FORTH).