Abstract
Anomaly detection is one of the innovative deep learning (DL) technologies being actively applied in various industrial fields. Among them, the supervised learning approach for anomaly detection requires a large amount of normal and abnormal data. However, in real industrial situations, obtaining abnormal data can be very expensive and difficult. Therefore, in this paper, we used just normal data in a semi-supervised learning approach. 72 geometric transformations of the input sample were added to train a DL network, and to classify the transformation type using the trained network. The proposed method was applied to detect mechanical part defects on an automobile industry production line. Because it is difficult to obtain enough abnormal data in the same region of interest (ROI), to improve classification performance, a method called outlier exposal is suggested, in which the normal data in another ROI is added as abnormal data for training. Abnormal data creates a uniform distribution in the output of the network so that the class of the prediction does not belong to any kind of normal data class. Abnormal data acts the anomaly detector to be generalized more and to increase the anomaly score for new abnormal data. Experimental results verified the proposed method improved performance compared to the conventional approach.
Similar content being viewed by others
References
Meshram, A., & Haas, C. (2017). Anomaly detection in industrial networks using machine learning: a roadmap. In Machine learning for cyber physical systems (pp. 65–72). Springer.
Sultani, W., Chen, C., & Shah, M. (2018). Real-world anomaly detection in surveillance videos. Paper presented at the Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 6479–6488).
Ukil, A., Bandyoapdhyay, S., Puri, C., & Pal, A. (2016). IoT healthcare analytics: The importance of anomaly detection. Paper presented at the 2016 IEEE 30th international conference on advanced information networking and applications (AINA) (pp. 994–997).
Chandola, V., Banerjee, A., & Kumar, V. (2009). Anomaly detection: A survey. ACM Computing Surveys (CSUR), 41(3), 1–58.
Chalapathy, R., Menon, A. K., & Chawla, S. (2018). Anomaly detection using one-class neural networks. ArXiv Preprint arXiv:1802.06360.
Kwon, B., Won, J., & Kang, D. (2015). Fast defect detection for various types of surfaces using random forest with VOV features. International Journal of Precision Engineering and Manufacturing, 16(5), 965–970.
Choi, E., & Kim, J. (2020). Deep learning based defect inspection using the intersection over minimum between search and abnormal regions. International Journal of Precision Engineering and Manufacturing, 21(4), 747–758.
Park, J., Kwon, B., Park, J., & Kang, D. (2016). Machine learning-based imaging system for surface defect inspection. International Journal of Precision Engineering and Manufacturing-Green Technology, 3(3), 303–310.
Park, J. K., An, W. H., & Kang, D. J. (2019). Convolutional neural network based surface inspection system for non-patterned welding defects. International Journal of Precision Engineering and Manufacturing, 20(3), 363–374.
Ringberg, H., Soule, A., Rexford, J., & Diot, C. (2007). Sensitivity of PCA for traffic anomaly detection. Paper presented at the Proceedings of the 2007 ACM SIGMETRICS international conference on measurement and modeling of computer systems (pp. 109–120).
Sakurada, M., & Yairi, T. (2014). Anomaly detection using autoencoders with nonlinear dimensionality reduction. Paper presented at the Proceedings of the MLSDA 2014 2nd Workshop on machine learning for sensory data analysis (pp. 4–11).
An, J., & Cho, S. (2015). Variational autoencoder based anomaly detection using reconstruction probability. Special Lecture on IE, 2(1), 1–18.
Ruff, L., Vandermeulen, R. A., Görnitz, N., Binder, A., Müller, E., Müller, K., & Kloft, M. (2019). Deep semi-supervised anomaly detection. ArXiv Preprint arXiv:1906.02694,
Ruff, L., Vandermeulen, R., Goernitz, N., Deecke, L., Siddiqui, S. A., Binder, A., Müller, E., & Kloft, M. (2018). Deep one-class classification. Paper presented at the international conference on machine learning (pp. 4393–4402).
Zagoruyko, S., & Komodakis, N. (2016). Wide residual networks. ArXiv Preprint arXiv:1605.07146
Goodfellow, I. J., Pouget-Abadie, J., Mirza, M., Xu, B., Warde-Farley, D., Ozair, S., Courville, A., & Bengio, Y. (2014). Generative adversarial networks. ArXiv Preprint arXiv:1406.2661.
Akcay, S., Atapour-Abarghouei, A., & Breckon, T. P. (2018). Ganomaly: Semi-supervised anomaly detection via adversarial training. Paper presented at the Asian conference on computer vision (pp. 622–637).
Deecke, L., Vandermeulen, R., Ruff, L., Mandt, S., & Kloft, M. (2018). Image anomaly detection with generative adversarial networks. Paper presented at the Joint European conference on machine learning and knowledge discovery in databases (pp. 3–17).
Golan, I., & El-Yaniv, R. (2018). Deep anomaly detection using geometric transformations. ArXiv Preprint arXiv:1805.10917.
Hendrycks, D., Mazeika, M., & Dietterich, T. (2018). Deep anomaly detection with outlier exposure. ArXiv Preprint arXiv:1812.04606.
He, K., Zhang, X., Ren, S., & Sun, J. (2016). Deep residual learning for image recognition. Paper presented at the Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 770–778).
Rubinstein, R. Y., & Kroese, D. P. (2013). The cross-entropy method: a unified approach to combinatorial optimization, Monte-Carlo simulation and machine learning. Springer.
Kingma, D. P., & Ba, J. (2014). Adam: A method for stochastic optimization. ArXiv Preprint arXiv:1412.6980.
Acknowledgements
This work was supported by the Center for Advanced Meta-Materials (CAMM) funded by the Ministry of Science and ICT as Global Frontier Project (CAMM- Nos. 2019M3A6B3031048 and 2014M3A6B3063700) and partly supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (No. 2021R1A2C1010057).
Author information
Authors and Affiliations
Contributions
Both authors read and approved the final manuscript.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Kwon, YW., Kang, DJ. Anomaly Detecting Geometric Transformation Network with Outlier Exposure Defect Inspection of Real Industrial Data. Int. J. Precis. Eng. Manuf. 24, 73–81 (2023). https://doi.org/10.1007/s12541-022-00736-w
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s12541-022-00736-w