Cross-domain Transfer of Defect Features in Technical Domains Based on Partial Target Data

##plugins.themes.bootstrap3.article.main##

##plugins.themes.bootstrap3.article.sidebar##

Published May 8, 2023
Tobias Schlagenhauf Tim Scheurenbrand

Abstract

A common challenge in real-world classification scenarios with sequentially appending target domain data is insufficient training datasets during the training phase. Therefore, conventional deep learning and transfer learning classifiers are not applicable especially when individual classes are not represented or are severely underrepresented at the outset. Domain Generalization approaches reach their limits when domain shifts become too large, making them occasionally unsuitable as well. In many (technical) domains, however, it is only the defect/ worn/ reject classes that are insufficiently represented, while the non-defect class is often available from the beginning. The proposed classification approach addresses such conditions. Following a contrastive learning approach, a CNN encoder is trained with a modified triplet loss function using two datasets: Besides the non-defective target domain class (= 1st dataset), a state-of-the-art labeled source domain dataset that contains highly related classes (e.g., a related manufacturing error or wear defect) but originates from a (highly) different domain (e.g., different product, material, or appearance) (= 2nd dataset) is utilized. The approach learns the classification features from the source domain dataset while at the same time learning the differences between the source and the target domain in a single training step, aiming to transfer the relevant features to the target domain. The classifier becomes sensitive to the classification features and – by architecture – robust against the highly domain-specific context. The approach is benchmarked in a technical and a non-technical domain and shows convincing classification results. In particular, it is shown that the domain generalization capabilities and classification results are improved by the proposed architecture, allowing for larger domain shifts between source and target domains.

Abstract 261 | PDF Downloads 296

##plugins.themes.bootstrap3.article.details##

Keywords

Domain Transfer; Domain Generalization

References
Ben-David, S., Blitzer, J., Crammer, K., Kulesza, A., Pereira, F., & Vaughan, J. W. (2010). A theory of learning from different domains. Machine Learning, 79(1–2), 151–175. https://doi.org/10.1007/s10994-009-5152-4
Blanchard, G., Lee, G., & Scott, C. (2011). Generalizing from Several Related Classification Tasks to a New Unlabeled Sample. https://papers.nips.cc/paper/2011/file/b571ecea16a9824023ee1af16897a582-Paper.pdf
Chen, T., Kornblith, S., Norouzi, M., & Hinton, G. (2020). A Simple Framework for Contrastive Learning of Visual Representations.
Chohan, M., Khan, A., Chohan, R., Katpar, S. H., & Mahar, M. S. (2020). Plant Disease Detection using Deep Learning. International Journal of Recent Technology and Engineering (IJRTE), 9(1), 909–914. https://doi.org/10.35940/ijrte.A2139.059120
Hamadache, M., Jung, J. H., Park, J., & Youn, B. D. (2019). A comprehensive review of artificial intelligence-based approaches for rolling element bearing PHM: shallow and deep learning. JMST Advances, 1(1–2), 125–151. https://doi.org/10.1007/s42791-019-0016-y
He, K., Fan, H., Wu, Y., Xie, S., & Girshick, R. (2019). Momentum Contrast for Unsupervised Visual Representation Learning.
He, Y., Shen, Z., & Cui, P. (2019). Towards Non-I.I.D. Image Classification: A Dataset and Baselines.
Huang, J., Guan, D., Xiao, A., & Lu, S. (2021). FSDR: Frequency Space Domain Randomization for Domain Generalization.
Hughes, David. P., & Salathe, M. (2015). An open access repository of images on plant health to enable the development of mobile disease diagnostics. https://data.mendeley.com/datasets/tywbtsjrjv/1
Jaiswal, A., Babu, A. R., Zadeh, M. Z., Banerjee, D., & Makedon, F. (2020). A Survey on Contrastive Self-supervised Learning. http://arxiv.org/abs/2011.00362
Kaiser, J.-P., Mitschke, N., Stricker, N., Heizmann, M., & Lanza, G. (2021). Konzept einer automatisierten und modularen Befundungsstation in der wandlungsfähigen Produktion. Zeitschrift Für Wirtschaftlichen Fabrikbetrieb, 116(5), 313–317. https://doi.org/10.1515/zwf-2021-0070
Kim, D., Wang, K., Sclaroff, S., & Saenko, K. (2022). A Broad Study of Pre-training for Domain Generalization and Adaptation.
Li, D., Yang, Y., Song, Y.-Z., & Hospedales, T. M. (2017). Deeper, Broader and Artier Domain Generalization.
Makerere AI Lab. (2020). Bean disease dataset. https://github.com/AI-Lab-Makerere/ibean/
Maqsood, M., Nazir, F., Khan, U., Aadil, F., Jamal, H., Mehmood, I., & Song, O. (2019). Transfer Learning Assisted Classification and Detection of Alzheimer’s Disease Stages Using 3D MRI Scans. Sensors, 19(11), 2645. https://doi.org/10.3390/s19112645
Pan, S. J., & Yang, Q. (2010). A Survey on Transfer Learning. IEEE Transactions on Knowledge and Data Engineering, 22(10), 1345–1359. https://doi.org/10.1109/TKDE.2009.191
Rombach, K., Michau, Dr. G., & Fink, Prof. Dr. O. (2022). Controlled Generation of Unseen Faults for Partial and Open-Partial Domain Adaptation.
Schlagenhauf, T. (2021). Ball Screw Drive Surface Defect Dataset for Classification (K. I. für Technologie, Ed.). Karlsruher Institut für Technologie (KIT) wbk Institute of Production Science. https://doi.org/10.5445/IR/1000133819
Schlagenhauf, T., Scheurenbrand, T., Hofmann, D., & Krasnikow, O. (2022). Analysis of the Visually Detectable Wear Progress on Ball Screws.
Severstal. (2020, December 8). Severstal: Steel Defect Detection (Severstal, Ed.). https://www.kaggle.com/c/severstal-steel-defect-detection/data
Shen, K., Jones, R., Kumar, A., Xie, S. M., HaoChen, J. Z., Ma, T., & Liang, P. (2022). Connect, Not Collapse: Explaining Contrastive Learning for Unsupervised Domain Adaptation. http://arxiv.org/abs/2204.00570
Siddiqui, M. W. (2015). Postharvest Biology and Technology of Horticultural Crops. Apple Academic Press.
Thota, M., & Leontidis, G. (2021). Contrastive Domain Adaptation.
Torralba, A., & Efros, A. A. (2011). Unbiased Look at Dataset Bias. CVPR 2011, 1521–1528. https://doi.org/10.1109/CVPR.2011.5995347
Wang, Y., Li, H., Chau, L., & Kot, A. C. (2021). Embracing the Dark Knowledge: Domain Generalization Using Regularized Knowledge Distillation. Proceedings of the 29th ACM International Conference on Multimedia, 2595–2604. https://doi.org/10.1145/3474085.3475434
Wang, Y., Li, H., & Kot, A. C. (2020). Heterogeneous Domain Generalization via Domain Mixup. https://doi.org/10.1109/ICASSP40776.2020.9053273
Xu, Z., Li, W., Niu, L., & Xu, D. (2014). Exploiting Low-Rank Structure from Latent Domains for Domain Generalization (pp. 628–643). https://doi.org/10.1007/978-3-319-10578-9_41
Xue, Y., Whitecross, K., & Mirzasoleiman, B. (2022). Investigating Why Contrastive Learning Benefits Robustness Against Label Noise.
Yang, C., Cheung, Y. M., Ding, J., Tan, K. C., Xue, B., & Zhang, M. (2022). Contrastive Learning Assisted-Alignment for Partial Domain Adaptation. IEEE Transactions on Neural Networks and Learning Systems. https://doi.org/10.1109/TNNLS.2022.3145034
Zhang, X., Cui, P., Xu, R., Zhou, L., He, Y., & Shen, Z. (2021). Deep Stable Learning for Out-Of-Distribution Generalization.
Zhang, X., Xu, Z., Xu, R., Liu, J., Cui, P., Wan, W., Sun, C., & Li, C. (2022). Towards Domain Generalization in Object Detection.
Zhou, K., Liu, Z., Qiao, Y., Xiang, T., & Loy, C. C. (2021). Domain Generalization: A Survey.
Zhou, K., Yang, Y., Hospedales, T., & Xiang, T. (2020). Deep Domain-Adversarial Image Generation for Domain Generalisation. Proceedings of the AAAI Conference on Artificial Intelligence, 34(07), 13025–13032. https://doi.org/10.1609/aaai.v34i07.7003
Section
Technical Papers