AI-driven methods link spectral fingerprints to structures and properties, providing a foundation for the unified inverse design of functional substances and delivering interpretable insights into universal spectrum–structure–property relationships.
Feng et al. (2025) studied this question.