PAH Marks the Spot: Digging for Buried Star Clusters in Nearby Galaxies

Identifying and characterizing embedded star formation is a key goal for the Physics at High Angular Resolution in Nearby GalaxieS (PHANGS) project. Using JWST and HST imaging we have identified hundreds of deeply embedded stellar clusters in a sample of 11 nearby star-forming galaxies using both a Zooniverse-based by-eye approach as well as machine learning. The main criterion defining `embedded' in this study is visible in JWST near-infrared imaging and invisible in HST UV/optical imaging. I will present several facets of this study: the similarities and differences between the samples generated via the by-eye and machine learning approaches; how the machine learning identifications depend on the convolutional neural networks employed; the distributions of the cluster profiles, ages, masses, and extinctions and how they compare to those of samples of exposed stellar clusters; and the spatial correlations between embedded clusters and the molecular gas as observed by ALMA.