PhaseMatcher: Autoregressive Phase-Set Identification with Spectral Decomposition

2026-10-05

AI Summary

This study addresses the challenge of incomplete phase identification in powder X-ray diffraction, where weak phase peaks are obscured by dominant signals. We propose a physics-guided autoregressive spectral decomposition framework that iteratively identifies phases through progressive decomposition and dynamic residual correction. To replace conventional scalar subtraction, we introduce a global contribution re-estimation mechanism based on raw observations and reference patterns, substantially improving residual estimation accuracy. Experimental evaluations on synthetic and controlled mixture datasets demonstrate that the proposed framework significantly outperforms existing baseline methods in complete phase set identification rates as well as in the accuracy of both contribution and residual estimations.

Abstract

Recovering complete phase sets from powder X-ray diffraction (PXRD) is challenging when weak-phase peaks overlap stronger signals. A natural strategy is to identify phases iteratively, removing the contribution of each identified phase from the observed pattern before predicting the next. However, even after a phase is correctly identified, misestimating its contribution can distort the residual and cause subsequent errors. We introduce PhaseMatcher, an autoregressive framework for complete phase-set identification with physics-guided spectral decomposition. After each phase prediction, PhaseMatcher re-estimates the contributions of all selected phases and the residual from the original observation and all selected reference patterns, accounting for physically plausible variation between reference patterns and the corresponding phase contributions in the observation. The resulting residual guides subsequent phase identification, while a separate stopping module determines when the phase set is complete. On synthetic mixtures and controlled mixtures constructed from measured single-phase patterns, PhaseMatcher improves complete-set identification over the evaluated baselines. On PhaseMix-135K, it also estimates contributions and residuals more accurately than scalar subtraction.

Problem

• Powder X-ray diffraction
• Phase identification
• Phase-set recovery
• Peak overlap
• Residual distortion

Innovation

• Autoregressive framework
• Spectral decomposition
• Phase-set identification
• Powder X-ray diffraction
• Residual estimation