Improving Multislice Electron Ptychography with a Generative Prior
Multislice electron ptychography (MEP) is an inverse imaging technique that computationally reconstructs the highest-resolution images of atomic crystal structures from diffraction patterns. Available algorithms often solve this inverse problem iteratively but are both time consuming and produce suboptimal solutions due to their ill-posed nature. We develop MEP-Diffusion, a diffusion model trained on a large database of crystal structures specifically for MEP to augment existing iterative solvers. MEP-Diffusion is easily integrated as a generative prior into existing reconstruction methods via Diffusion Posterior Sampling (DPS). We find that this hybrid approach greatly enhances the quality of the reconstructed 3D volumes, achieving a 90.50% improvement in SSIM over existing methods.
Code (0)
등록된 구현이 없습니다.
Similar Papers 제목 키워드 기반
Autonomous Algorithm Discovery for Ptychography via Evolutionary LLM Reasoning
Ptychography is a computational imaging technique widely used for high-resolution materials characterization, but high-quality reconstructions often require the use of regularization functions that largely remain manuall…
Code GenerationNear-Isotropic Sub-Ångstrom 3D Resolution Phase Contrast Imaging Achieved by End-to-End Ptychographic Electron Tomography
Three-dimensional atomic resolution imaging using transmission electron microscopes is a unique capability that requires challenging experiments. Linear electron tomography methods are limited by the missing wedge effect…
Electron TomographyOvercoming information reduced data and experimentally uncertain parameters in ptychography with regularized optimization
The overdetermination of the mathematical problem underlying ptychography is reduced by a host of experimentally more desirable settings. Furthermore, reconstruction of the sample-induced phase shift is typically limited…
Exploring Generative Adversarial Networks for Image-to-Image Translation in STEM Simulation
The use of accurate scanning transmission electron microscopy (STEM) image simulation methods require large computation times that can make their use infeasible for the simulation of many images. Other simulation methods…
Deep LearningGenerative Adversarial NetworkImage-to-Image Translationregression+1Deep Ptych: Subsampled Fourier Ptychography using Generative Priors
This paper proposes a novel framework to regularize the highly ill-posed and non-linear Fourier ptychography problem using generative models. We demonstrate experimentally that our proposed algorithm, Deep Ptych, outperf…