skip to main content
DOE PAGES title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: A Hybrid Deep Learning Architecture for Classification of Microscopic Damage on NIF Laser Optics

Abstract

Accurately classifying microscopic damage helps automate the repair and recycling of National Ignition Facility optics and informs the study of damage initiation and growth. This complex 12-class problem previously required human experts to distinguish and label the various damage morphologies. Finding image analysis methods to extract and calculate distinguishing features would be time consuming and challenging, so we sought to automate this task by using convolutional neural networks (CNNs) pretrained on the ImageNet database to take advantage of its automated feature discovery and extraction. We compared three model architectures on this dataset and found the one with highest overall accuracy, 99.17%, was a novel hybrid architecture, one in which we removed the final decision-making layer of the deep learner and replaced it with an ensemble of decision trees (EDT). This combines the power of feature extraction by CNNs with the decision-making strength of EDT. The accuracy of the hybrid architecture over the deep learning alone is shown to be significantly improved. Furthermore, we applied this novel hybrid architecture to an entirely different dataset, one containing images of repaired damage sites, and improved on the previously published findings, also with a demonstrably significant increase in accuracy over using the deep learnermore » alone.« less

Authors:
 [1];  [1];  [2]
  1. Lawrence Livermore National Lab. (LLNL), Livermore, CA (United States)
  2. Sandia National Lab. (SNL-CA), Livermore, CA (United States)
Publication Date:
Research Org.:
Lawrence Livermore National Lab. (LLNL), Livermore, CA (United States)
Sponsoring Org.:
USDOE National Nuclear Security Administration (NNSA)
OSTI Identifier:
1651187
Alternate Identifier(s):
OSTI ID: 1568903
Report Number(s):
LLNL-JRNL-764783
Journal ID: ISSN 1932-1864; 954634
Grant/Contract Number:  
AC52-07NA27344; DE‐AC52‐07NA27344
Resource Type:
Accepted Manuscript
Journal Name:
Statistical Analysis and Data Mining
Additional Journal Information:
Journal Volume: 12; Journal Issue: 6; Journal ID: ISSN 1932-1864
Publisher:
Wiley
Country of Publication:
United States
Language:
English
Subject:
Lasers; 47 OTHER INSTRUMENTATION

Citation Formats

Amorin, C., Kegelmeyer, L., and Kegelmeyer, P.. A Hybrid Deep Learning Architecture for Classification of Microscopic Damage on NIF Laser Optics. United States: N. p., 2019. Web. https://doi.org/10.1002/sam.11437.
Amorin, C., Kegelmeyer, L., & Kegelmeyer, P.. A Hybrid Deep Learning Architecture for Classification of Microscopic Damage on NIF Laser Optics. United States. https://doi.org/10.1002/sam.11437
Amorin, C., Kegelmeyer, L., and Kegelmeyer, P.. Mon . "A Hybrid Deep Learning Architecture for Classification of Microscopic Damage on NIF Laser Optics". United States. https://doi.org/10.1002/sam.11437. https://www.osti.gov/servlets/purl/1651187.
@article{osti_1651187,
title = {A Hybrid Deep Learning Architecture for Classification of Microscopic Damage on NIF Laser Optics},
author = {Amorin, C. and Kegelmeyer, L. and Kegelmeyer, P.},
abstractNote = {Accurately classifying microscopic damage helps automate the repair and recycling of National Ignition Facility optics and informs the study of damage initiation and growth. This complex 12-class problem previously required human experts to distinguish and label the various damage morphologies. Finding image analysis methods to extract and calculate distinguishing features would be time consuming and challenging, so we sought to automate this task by using convolutional neural networks (CNNs) pretrained on the ImageNet database to take advantage of its automated feature discovery and extraction. We compared three model architectures on this dataset and found the one with highest overall accuracy, 99.17%, was a novel hybrid architecture, one in which we removed the final decision-making layer of the deep learner and replaced it with an ensemble of decision trees (EDT). This combines the power of feature extraction by CNNs with the decision-making strength of EDT. The accuracy of the hybrid architecture over the deep learning alone is shown to be significantly improved. Furthermore, we applied this novel hybrid architecture to an entirely different dataset, one containing images of repaired damage sites, and improved on the previously published findings, also with a demonstrably significant increase in accuracy over using the deep learner alone.},
doi = {10.1002/sam.11437},
journal = {Statistical Analysis and Data Mining},
number = 6,
volume = 12,
place = {United States},
year = {2019},
month = {9}
}

Journal Article:
Free Publicly Available Full Text
Publisher's Version of Record

Citation Metrics:
Cited by: 1 work
Citation information provided by
Web of Science

Save / Share:

Works referenced in this record:

ImageNet: A large-scale hierarchical image database
conference, June 2009

  • Deng, Jia; Dong, Wei; Socher, Richard
  • 2009 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops (CVPR Workshops), 2009 IEEE Conference on Computer Vision and Pattern Recognition
  • DOI: 10.1109/CVPR.2009.5206848

Deep Residual Learning for Image Recognition
conference, June 2016

  • He, Kaiming; Zhang, Xiangyu; Ren, Shaoqing
  • 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)
  • DOI: 10.1109/CVPR.2016.90

A Survey on Transfer Learning
journal, October 2010

  • Pan, Sinno Jialin; Yang, Qiang
  • IEEE Transactions on Knowledge and Data Engineering, Vol. 22, Issue 10
  • DOI: 10.1109/TKDE.2009.191

Optics Recycle Loop Strategy for NIF Operations above UV Laser-Induced Damage Threshold
journal, February 2016

  • Spaeth, M. L.; Wegner, P. J.; Suratwala, T. I.
  • Fusion Science and Technology, Vol. 69, Issue 1
  • DOI: 10.13182/FST15-119

Growth model for laser-induced damage on the exit surface of fused silica under UV, ns laser irradiation
journal, January 2014

  • Negres, Raluca A.; Cross, David A.; Liao, Zhi M.
  • Optics Express, Vol. 22, Issue 4
  • DOI: 10.1364/OE.22.003824

Growth behavior of laser-induced damage on fused silica optics under UV, ns laser irradiation
journal, January 2010

  • Negres, Raluca A.; Norton, Mary A.; Cross, David A.
  • Optics Express, Vol. 18, Issue 19
  • DOI: 10.1364/OE.18.019966

Automated optics inspection analysis for NIF
journal, December 2012


Neural Random Forests
journal, June 2018


Neural Decision Forests for Semantic Image Labelling
conference, June 2014

  • Rota Bulo, Samuel; Kontschieder, Peter
  • 2014 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)
  • DOI: 10.1109/CVPR.2014.18

Factors of Transferability for a Generic ConvNet Representation
journal, September 2016

  • Azizpour, Hossein; Razavian, Ali Sharif; Sullivan, Josephine
  • IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 38, Issue 9
  • DOI: 10.1109/TPAMI.2015.2500224

A Comparison of Decision Tree Ensemble Creation Techniques
journal, January 2007

  • Banfield, Robert E.; Hall, Lawrence O.; Bowyer, Kevin W.
  • IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 29, Issue 1
  • DOI: 10.1109/TPAMI.2007.250609

Mitigation of laser damage on National Ignition Facility optics in volume production
conference, December 2013

  • Folta, James; Nostrand, Mike; Honig, John
  • SPIE Laser Damage, SPIE Proceedings
  • DOI: 10.1117/12.2030475

Going deeper with convolutions
conference, June 2015


Caffe: Convolutional Architecture for Fast Feature Embedding
conference, January 2014

  • Jia, Yangqing; Shelhamer, Evan; Donahue, Jeff
  • Proceedings of the ACM International Conference on Multimedia - MM '14
  • DOI: 10.1145/2647868.2654889