Improving robustness for model discerning synthesis process of uranium oxide with unsupervised domain adaptation
The quantitative characterization of surface structures captured in scanning electron microscopy (SEM) images has proven to be effective for discerning provenance of an unknown nuclear material. Recently, many works have taken advantage of the powerful performance of convolutional neural networks (C...
Main Authors: | , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
Frontiers Media S.A.
2023-10-01
|
Series: | Frontiers in Nuclear Engineering |
Subjects: | |
Online Access: | https://www.frontiersin.org/articles/10.3389/fnuen.2023.1230052/full |
_version_ | 1797650172844441600 |
---|---|
author | Cuong Ly Cuong Ly Cuong Ly Cody Nizinski Cody Nizinski Alex Hagen Luther W McDonald Tolga Tasdizen Tolga Tasdizen |
author_facet | Cuong Ly Cuong Ly Cuong Ly Cody Nizinski Cody Nizinski Alex Hagen Luther W McDonald Tolga Tasdizen Tolga Tasdizen |
author_sort | Cuong Ly |
collection | DOAJ |
description | The quantitative characterization of surface structures captured in scanning electron microscopy (SEM) images has proven to be effective for discerning provenance of an unknown nuclear material. Recently, many works have taken advantage of the powerful performance of convolutional neural networks (CNNs) to provide faster and more consistent characterization of surface structures. However, one inherent limitation of CNNs is their degradation in performance when encountering discrepancy between training and test datasets, which limits their use widely. The common discrepancy in an SEM image dataset occurs at low-level image information due to user-bias in selecting acquisition parameters and microscopes from different manufacturers. Therefore, in this study, we present a domain adaptation framework to improve robustness of CNNs against the discrepancy in low-level image information. Furthermore, our proposed approach makes use of only unlabeled test samples to adapt a pretrained model, which is more suitable for nuclear forensics application for which obtaining both training and test datasets simultaneously is a challenge due to data sensitivity. Through extensive experiments, we demonstrate that our proposed approach effectively improves the performance of a model by at least 18% when encountering domain discrepancy, and can be deployed in many CNN architectures. |
first_indexed | 2024-03-11T15:56:26Z |
format | Article |
id | doaj.art-ed149e3821d04820b5731a767410abe1 |
institution | Directory Open Access Journal |
issn | 2813-3412 |
language | English |
last_indexed | 2024-03-11T15:56:26Z |
publishDate | 2023-10-01 |
publisher | Frontiers Media S.A. |
record_format | Article |
series | Frontiers in Nuclear Engineering |
spelling | doaj.art-ed149e3821d04820b5731a767410abe12023-10-25T10:30:49ZengFrontiers Media S.A.Frontiers in Nuclear Engineering2813-34122023-10-01210.3389/fnuen.2023.12300521230052Improving robustness for model discerning synthesis process of uranium oxide with unsupervised domain adaptationCuong Ly0Cuong Ly1Cuong Ly2Cody Nizinski3Cody Nizinski4Alex Hagen5Luther W McDonald6Tolga Tasdizen7Tolga Tasdizen8Department of Electrical and Computer Engineering, College of Engineering, University of Utah, Salt Lake City, UT, United StatesPacific Northwest National Laboratory (DOE), Richland, UT, United StatesScientific Computing and Imaging Institute, College of Engineering, University of Utah, Salt Lake City, UT, United StatesPacific Northwest National Laboratory (DOE), Richland, UT, United StatesDepartment of Civil and Environmental Engineering, College of Engineering, University of Utah, Salt Lake City, UT, United StatesPacific Northwest National Laboratory (DOE), Richland, UT, United StatesDepartment of Civil and Environmental Engineering, College of Engineering, University of Utah, Salt Lake City, UT, United StatesDepartment of Electrical and Computer Engineering, College of Engineering, University of Utah, Salt Lake City, UT, United StatesScientific Computing and Imaging Institute, College of Engineering, University of Utah, Salt Lake City, UT, United StatesThe quantitative characterization of surface structures captured in scanning electron microscopy (SEM) images has proven to be effective for discerning provenance of an unknown nuclear material. Recently, many works have taken advantage of the powerful performance of convolutional neural networks (CNNs) to provide faster and more consistent characterization of surface structures. However, one inherent limitation of CNNs is their degradation in performance when encountering discrepancy between training and test datasets, which limits their use widely. The common discrepancy in an SEM image dataset occurs at low-level image information due to user-bias in selecting acquisition parameters and microscopes from different manufacturers. Therefore, in this study, we present a domain adaptation framework to improve robustness of CNNs against the discrepancy in low-level image information. Furthermore, our proposed approach makes use of only unlabeled test samples to adapt a pretrained model, which is more suitable for nuclear forensics application for which obtaining both training and test datasets simultaneously is a challenge due to data sensitivity. Through extensive experiments, we demonstrate that our proposed approach effectively improves the performance of a model by at least 18% when encountering domain discrepancy, and can be deployed in many CNN architectures.https://www.frontiersin.org/articles/10.3389/fnuen.2023.1230052/fullnuclear forensicsmachine learningconvolutional neural networksdomain adaptationscanning electron microscopy |
spellingShingle | Cuong Ly Cuong Ly Cuong Ly Cody Nizinski Cody Nizinski Alex Hagen Luther W McDonald Tolga Tasdizen Tolga Tasdizen Improving robustness for model discerning synthesis process of uranium oxide with unsupervised domain adaptation Frontiers in Nuclear Engineering nuclear forensics machine learning convolutional neural networks domain adaptation scanning electron microscopy |
title | Improving robustness for model discerning synthesis process of uranium oxide with unsupervised domain adaptation |
title_full | Improving robustness for model discerning synthesis process of uranium oxide with unsupervised domain adaptation |
title_fullStr | Improving robustness for model discerning synthesis process of uranium oxide with unsupervised domain adaptation |
title_full_unstemmed | Improving robustness for model discerning synthesis process of uranium oxide with unsupervised domain adaptation |
title_short | Improving robustness for model discerning synthesis process of uranium oxide with unsupervised domain adaptation |
title_sort | improving robustness for model discerning synthesis process of uranium oxide with unsupervised domain adaptation |
topic | nuclear forensics machine learning convolutional neural networks domain adaptation scanning electron microscopy |
url | https://www.frontiersin.org/articles/10.3389/fnuen.2023.1230052/full |
work_keys_str_mv | AT cuongly improvingrobustnessformodeldiscerningsynthesisprocessofuraniumoxidewithunsuperviseddomainadaptation AT cuongly improvingrobustnessformodeldiscerningsynthesisprocessofuraniumoxidewithunsuperviseddomainadaptation AT cuongly improvingrobustnessformodeldiscerningsynthesisprocessofuraniumoxidewithunsuperviseddomainadaptation AT codynizinski improvingrobustnessformodeldiscerningsynthesisprocessofuraniumoxidewithunsuperviseddomainadaptation AT codynizinski improvingrobustnessformodeldiscerningsynthesisprocessofuraniumoxidewithunsuperviseddomainadaptation AT alexhagen improvingrobustnessformodeldiscerningsynthesisprocessofuraniumoxidewithunsuperviseddomainadaptation AT lutherwmcdonald improvingrobustnessformodeldiscerningsynthesisprocessofuraniumoxidewithunsuperviseddomainadaptation AT tolgatasdizen improvingrobustnessformodeldiscerningsynthesisprocessofuraniumoxidewithunsuperviseddomainadaptation AT tolgatasdizen improvingrobustnessformodeldiscerningsynthesisprocessofuraniumoxidewithunsuperviseddomainadaptation |