Inferring low-dimensional microstructure representations using convolutional neural networks

Abstract

We apply recent advances in machine learning and computer vision to a central problem in materials informatics: The statistical representation of microstructural images. We use activations in a pre-trained convolutional neural network to provide a high-dimensional characterization of a set of synthetic microstructural images. Next, we use manifold learning to obtain a low-dimensional embedding of this statistical characterization. We show that the low-dimensional embedding extracts the parameters used to generate the images. According to a variety of metrics, the convolutional neural network method yields dramatically better embeddings than the analogous method derived from two-point correlations alone.

14 Pages, 12 Figures

References in corpus (4)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• Modeling Heterogeneous Materials via Two-Point Correlation Functions: I. Basic Principles
• Geometrical Ambiguity of Pair Statistics. II. Heterogeneous Media
• Transformation Properties of Learned Visual Representations

Cited by in corpus (16)

• Overview: Computer vision and machine learning for microstructural characterization and analysis
• Microstructural Materials Design via Deep Adversarial Learning Methodology
• Geometric deep learning for computational mechanics Part I: Anisotropic Hyperelasticity
• Microstructure synthesis using style-based generative adversarial network
• The Materials Simulation Toolkit for Machine Learning (MAST-ML): an automated open source toolkit to accelerate data-driven materials research
• Predicting permeability via statistical learning on higher-order microstructural information
• Artificial intelligence approaches for materials-by-design of energetic materials: state-of-the-art, challenges, and future directions
• Precise algorithms to compute surface correlation functions of two-phase heterogeneous media and their applications
• Two-stage 2D-to-3D reconstruction of realistic microstructures: Implementation and numerical validation by effective properties
• A Transfer Learning Approach for Microstructure Reconstruction and Structure-property Predictions
• Machine learning electron correlation in a disordered medium
• Surrogate- and invariance-boosted contrastive learning for data-scarce applications in science
• Predicting Peak Stresses In Microstructured Materials Using Convolutional Encoder-Decoder Learning
• Building Data-driven Models with Microstructural Images: Generalization and Interpretability
• Predicting the mechanical response of oligocrystals with deep learning
• An efficient optimization based microstructure reconstruction approach with multiple loss functions