Chinese AI SpecCLIP Unifies Stellar Data

A Chinese research team, including scientists from the China===Chinese Academy of Sciences and the University of the Chinese Academy of Sciences, has developed an artificial intelligence model named SpecCLIP. This AI acts as a 'translator' for stellar spectral data, enabling the unified analysis of information collected by different telescopes, such as China's LAMOST and Europe's Gaia (spacecraft) satellite. Previously, varying methods, resolutions, and wavelength ranges made direct comparison and large-scale analysis difficult. SpecCLIP uses concepts similar to large language models and contrastive learning to establish intrinsic connections between disparate datasets. This breakthrough significantly accelerates research into the Milky Way's formation and evolution, and improves the efficiency of searching for potentially habitable planets by accurately characterizing planet-hosting stars. The study was published in The Astrophysical Journal and reported by Science and Technology Daily.