Work

The NEMESIS general young stellar object catalogue. I. Supervised classification with deep learning methods

Year: 2026

Type: article

Abstract: In modern large-scale photometric surveys, young stellar objects (YSOs) are mostly identified by the infrared excess seen in their spectral energy distributions (SEDs). To this end, YSO identification and classification were mostly based on certain colours specific to a dataset, resulting in wavelength-specific methods and catalogues. However, a uniformly classified YSO catalogue can be constructed by using all available data and applying the same identification method to heterogeneous literature samples, thereby reducing systematic differences between catalogues and enabling more consistent studies of YSO populations and star formation. Our primary goal is to develop a method that takes advantage of the heterogeneous data available in the VizieR database and to provide a tool that can be easily used to identify YSOs with a high accuracy. We also aim to create the largest homogeneous catalogue of YSOs to date. We used the CDS VizieR and NASA/IPAC Infrared Science Archive (IRSA) databases to collect data for a large number of celestial objects of various types. The data collected include SEDs, photometric light curves, and direct imaging data, and they were used to build different image representations for our list of sources. These image representations were used to teach deep learning algorithms to differentiate between YSOs and all other types of sources. Established architectures from the library and custom-built convolutional neural networks (CNNs) were applied to create an ensemble of binary classifiers specialised in identifying YSOs against other types of astrophysical sources. PyTorch The resulting method is able to automatically identify YSO candidates with an accuracy higher than 90%, while contamination is identified with an accuracy higher than 99%. We applied our classifier to numerous YSO catalogues from the literature to confirm the youth of the stars listed there and to create the largest catalogue of homogeneously identified potential YSOs ever, the NEMESIS General YSO (NGYSO) catalogue, which contains 274,408 unique high-reliability YSOs. Their positions on the colour-magnitude diagrams and their distribution relative to the local Galactic structures were also used to test their young nature.

Source: Astronomy and Astrophysics

DOI: 10.1051/0004-6361/202660312