TY - GEN
T1 - Image-based process monitoring via adversarial autoencoder with applications to rolling defect detection
AU - Yan, Hao
AU - Yeh, Huai Ming
AU - Sergin, Nurettin
N1 - Funding Information:
This work is funded by NSF DMS-183036. Please address all correspondence to Professor Hao Yan, HaoYan@asu.edu
Funding Information:
This work is funded by NSF DMS-183036
Publisher Copyright:
© 2019 IEEE.
PY - 2019/8
Y1 - 2019/8
N2 - Image-based process monitoring has recently attracted increasing attention due to the advancement of the sensing technologies. However, existing process monitoring methods fail to fully utilize the spatial information of images due to their complex characteristics including the high-dimensionality and complex spatial structures. Recent advancements in unsupervised deep models such as generative adversarial networks (GAN) and adversarial autoencoders (AAE) has enabled to learn the complex spatial structures automatically. Inspired by this advancement, we propose an anomaly detection framework based on the AAE for unsupervised anomaly detection for images. AAE combines the power of GAN with the variational autoencoder, which serves as a nonlinear dimension reduction technique. Based on this, we propose a monitoring statistic efficiently capturing the change of the data. The performance of the proposed AAE-based anomaly detection algorithm is validated through a simulation study and real case study for rolling defect detection.
AB - Image-based process monitoring has recently attracted increasing attention due to the advancement of the sensing technologies. However, existing process monitoring methods fail to fully utilize the spatial information of images due to their complex characteristics including the high-dimensionality and complex spatial structures. Recent advancements in unsupervised deep models such as generative adversarial networks (GAN) and adversarial autoencoders (AAE) has enabled to learn the complex spatial structures automatically. Inspired by this advancement, we propose an anomaly detection framework based on the AAE for unsupervised anomaly detection for images. AAE combines the power of GAN with the variational autoencoder, which serves as a nonlinear dimension reduction technique. Based on this, we propose a monitoring statistic efficiently capturing the change of the data. The performance of the proposed AAE-based anomaly detection algorithm is validated through a simulation study and real case study for rolling defect detection.
KW - Adversarial Autoencoder
KW - Deep Generative Models
KW - Profile Monitoring
KW - Statistical Process Control
UR - http://www.scopus.com/inward/record.url?scp=85072967019&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85072967019&partnerID=8YFLogxK
U2 - 10.1109/COASE.2019.8843313
DO - 10.1109/COASE.2019.8843313
M3 - Conference contribution
AN - SCOPUS:85072967019
T3 - IEEE International Conference on Automation Science and Engineering
SP - 311
EP - 316
BT - 2019 IEEE 15th International Conference on Automation Science and Engineering, CASE 2019
PB - IEEE Computer Society
T2 - 15th IEEE International Conference on Automation Science and Engineering, CASE 2019
Y2 - 22 August 2019 through 26 August 2019
ER -