Deep-Learning Based Microstructure Reconstruction and Generation

Author(s) Information

Cameron J. Maloney, California Polytechnic State University, San Luis ObispoFollow
Lucas Taliaferro, California Polytechnic State University, San Luis ObispoFollow
Nina St. John, California Polytechnic State University, San Luis ObispoFollow

College - Author 1

College of Engineering

Department - Author 1

Computer Science Department

College - Author 2

College of Engineering

Department - Author 2

Biomedical Engineering Department

College - Author 3

College of Engineering

Department - Author 3

Computer Engineering Department

Advisor

Jonathan Ventura, College of Engineering, Computer Science and Software Engineering Department; Long Wang, College of Engineering, Civil and Environmental Engineering Department

Funding Source

Noyce School of Applied Computing

Date

10-2025

Abstract/Summary

Microscopic imaging is essential to characterize multi-scale material behavior and understanding structure-property relationships. Recently our group developed a deep learning approach based on a Generative Adversarial Network (GAN) to reconstruct and artificially generate microstructures of strain-sensing nanomaterial networks based on microscope imagery. In this SURP project we would like to evaluate an alternative approach called diffusion to see if we can improve the quality of our results. Furthermore, we aim to test our approaches on a wider variety of materials, which will have different microstructures, to evaluate how versatile our models are.

October 1, 2025.