CMB-ML: A Cosmic Microwave Background Dataset for the Oldest Possible Computer Vision Task

James Amato, Yunan Xie, Leonel Medina-Varela, Ammar Aljerwi, Adam McCutcheon, T. Seth Rippentrop, Kristian Gonzalez, Jacques Delabrouille, Mustapha Ishak, Nicholas Ruozzi; Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), 2025, pp. 9418-9430

The Cosmic Microwave Background (CMB) radiation is a pillar of modern cosmology that gives rise to a better understanding of the fundamental parameters of the universe. While the astrophysics community has developed computational methods to extract this signal from data, these methods have limited scalability, and several groups have proposed the adoption of computer vision based models for CMB signal extraction. However, these diverse models are difficult to compare: the underlying datasets and evaluations are inconsistent and have not been made publicly available. We propose CMB-ML, a dataset and library that integrates dataset creation, model inference, and result evaluation into a pipeline to fill this gap and to make the problem accessible to researchers outside of cosmology. The library and links for data are available at github.com/CMB-ML/cmb-ml.