A New Cosmic-Ray Rejection Routine for HST WFC3/UVIS via Label-free Training of deepCR

Abstract

Abstract deepCR is a deep-learning-based cosmic-ray (CR) rejection framework originally presented by Zhang & Bloom. The original approach requires a dedicated training set that consists of multiple frames of the same fields, enabling automatic CR labeling through comparison with their median coadds. Here, we present a novel training approach that circumvents the need for a dedicated training set, but instead utilizes dark frames and the science images requiring CR removal themselves. During training, CRs present in dark frames are added to the science images, which the network is then trained to identify. In turn, the trained deepCR model can then be applied to identify CRs originally present in the science images. Using this approach, we present a new deepCR model trained on a diverse set of Hubble Space Telescope images taken from resolved galaxies in the Local Group, which is universally applicable across all WFC3/UVIS filters. We introduce a robust approach to determining the threshold for generating binary cosmic-ray masks from predictions from deepCR probability maps. When applied to the Panchromatic Hubble Andromeda Southern Treasury survey, our new deepCR model added ∼7% of good-quality stars that exhibit distinct features in their color–magnitude diagrams.