Quantitative X-Ray Microanalysis with Monte Carlo Simulations and Deep Learning************

利用蒙特卡罗模拟和深度学习进行定量 X 射线微量分析************

• 批准号: 537609-2018
• 负责人: GAUVIN, Raynald
• 金额: $ 1.82万
• 依托单位: McGill University
• 依托单位国家: 加拿大
• 项目类别: Engage Grants Program
• 财政年份: 2018
• 资助国家: 加拿大
• 起止时间: 2018-01-01 至 2019-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=651631
• 关键词: Quantitative; Ray; Microanalysis; Monte; Carlo

Convolutional neural networks (CNN) are currently implemented in a variety of image processing techniques. However, in order for them to be applied to digital signal processing, specifically X-ray spectra analysis, large and diverse training sets must be generated to cover a wide range of possible characteristics. To reach a diversity of materials science applications, training sets tailored to each specific problem must be appropriately compiled. The overall objective of the research project is to implement Monte Carlo X-ray simulations as training sets for CNNs in the ORS Dragonfly c software in order to provide a tool for X-ray spectra analysis that is efficient and accurate, and can cover a large variety of material science research based problems.

卷积神经网络（CNN）目前在各种图像处理技术中实现。然而，为了将它们应用于数字信号处理，特别是X射线光谱分析，必须生成大量不同的训练集，以覆盖广泛的可能特征。为了实现材料科学应用的多样性，必须适当地编译针对每个特定问题的训练集。该研究项目的总体目标是在ORS Dragonfly c软件中实现Monte Carlo X射线模拟作为CNN的训练集，以便为X射线光谱分析提供一种高效准确的工具，并且可以覆盖各种材料科学研究问题。