|About this Abstract
||Materials Science & Technology 2019
||Late News Poster Session
||P2-19: Predicting Compressive Strength of Consolidated Solids Using Computer Vision and Deep Learning
||T. Yong Han, Brian Gallagher, Donald Loveland, Matthew Rever, T. Nathan Mundhenk, Emily Robertson
|On-Site Speaker (Planned)
||T. Yong Han
We explore the application of computer vision and machine learning (ML) techniques to quantify material properties (e.g., compressive strength) based on SEM images of material microstructure. We show that it is possible to train ML models to predict materials performance based on SEM images alone, demonstrating this capability on the real-world problem of predicting uniaxially compressed peak stress of consolidated molecular solids samples. We compared two complementary approaches to this problem: (1) a traditional ML approach, random forest (RF), using state-of-the-art computer vision features and (2) an end-to-end deep learning (DL) approach, where features are learned automatically from raw images. We demonstrate the complementarity of these approaches, showing that RF performs best in the “small data” regime in which many real-world scientific applications reside, whereas DL outpaces RF in the “big data” regime, where abundant training samples are available.