Scholarlink

Abstract <p>An approach is proposed that can allow ab initio solution of the phase problem in X-ray diffraction studies. A synthetic diffraction data generator has been created that can be used to solve applied problems using machine learning. An automated container is presented that allows reproducible experiments aimed at solving the problem within the framework of the proposed approach. The FFT_UNet and XRD_Transformer models were developed, taking into account the physical specifics of the problem, and a comparative analysis and interpretation of their work on real data were carried out.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Comparison of Deep Learning Methods for Solving the Phase Problem in X-ray Diffraction Analysis

  • 期刊论文

  • 发表日期：

    2025年12月1日

查看全文
  • A. I. Khaibrakhmanov,
  • A. O. Dmitrienko

摘要

Abstract

An approach is proposed that can allow ab initio solution of the phase problem in X-ray diffraction studies. A synthetic diffraction data generator has been created that can be used to solve applied problems using machine learning. An automated container is presented that allows reproducible experiments aimed at solving the problem within the framework of the proposed approach. The FFT_UNet and XRD_Transformer models were developed, taking into account the physical specifics of the problem, and a comparative analysis and interpretation of their work on real data were carried out.