文献类型：期刊文献

英文题名：STEMDiff: A Wavelet-Enhanced Diffusion Model for Physics-Informed STEM Image Generation

作者：Bao, Yihui[1];Lu, Xinyi[1];Xia, Yanyan[1];Ye, Zhencheng[1];Chen, Houyang[2,3]

机构：[1]East China Univ Sci & Technol, Sch Informat Sci & Engn, Shanghai 200237, Peoples R China;[2]Chinese Acad Sci, Chongqing Inst Green & Intelligent Technol, Chongqing 400714, Peoples R China;[3]Univ Chinese Acad Sci, Chongqing Sch, Chongqing 400714, Peoples R China

年份：2025

卷号：12

期号：41

外文期刊名：ADVANCED SCIENCE

收录：;EI(收录号：20254119290704);WOS:【SCI-EXPANDED(收录号:WOS:001581810300001)】；

基金：This work is supported by the National Key Research and Development Program of China (2023YFB3307800), the Postdoctoral Program of Natural Science Foundation of Chongqing (CSTB2023NSCQ-BHX0159), the National Natural Science Foundation of China (62273149, 62373153), the Natural Science Foundation of Chongqing (CSTB2023NSCQ-MSX0045), the Startup Foundation of Chongqing Institute of Green and Intelligent Technology, Chinese Academy of Sciences. The authors acknowledge Hzwtech (www.hzwtech.com) for providing HPC resources that have contributed to the research results reported within this paper.

语种：英文

外文关键词：diffusion model; materials design; STEM image generation; STEM; STEMDiff

摘要：Machine learning has emerged as a powerful tool for analyzing scanning transmission electron microscopy (STEM) images, yet its widespread application remains constrained by the scarcity of annotated training data. While deep generative models offer a promising solution, they typically struggle to reproduce the complex high-frequency components that define experimental STEM images. Here, STEMDiff, a conditional diffusion model that transforms simple binary labels derived from crystal structures into realistic STEM images through a physical information embedding strategy, is proposed. By developing a novel Discrete Wavelet Transform-based skip-connection architecture, the high-frequency bias inherent in diffusion models are addressed, enabling the preservation of experimental noise characteristics. This approach generates images that are quantitatively nearly indistinguishable from experimental data (17 fold improvement over previous methods) while retaining ground truth structural information. Fully convolutional networks trained exclusively on these synthetic images achieve high-precision atomic column detection in experimental STEM images of WSe2 and graphene, despite the presence of substantial background noise and contamination. This approach effectively eliminates the need for laborious manual annotation, providing a scalable solution to the data bottleneck in STEM image analysis. The principles underlying STEMDiff can extend to other scientific imaging modalities, accelerating advancements in materials design for water treatment.