Abstract
The rapid renewal of industrial production lines is accompanied by the emergence of many unseen new defects that are difficult to detect, which may lead to large economic losses. Existing methods focus on the recognition of seen defects, but are powerless against unseen defects. The recognition of unseen defects is a challenging task and has not been widely explored. To our knowledge, we are the first to raise the issue of unseen defect recognition. To tackle this issue, we design an unseen defect generative adversarial network (UDGAN) model. The UDGAN not only explores the similarity between seen and unseen defects by extracting the latent distribution of the images but also represents the difference between these two defects considering the distribution interval. To enhance the understanding of the latent distribution of seen and generated images, we investigate their connection through mutual information optimization, giving some correlation between seen defects and generated images. Meanwhile, the distance between the mean of the seen and unseen defect distributions is expanded to optimize the extraction of latent distributions so that the types of defects in the generated and seen images are as different as possible. Experimental results on the magnetic tile defect dataset show that UDGAN achieves significant improvements over state-of-the-art methods.
Supported by the National Nature Science Foundation of China (U1803262).
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Arjovsky, M., Chintala, S., Bottou, L.: Wasserstein GAN (2017)
Belghazi, M.I., et al.: Mutual information neural estimation. In: International Conference on Machine Learning, pp. 531–540. PMLR (2018)
Chen, X., Duan, Y., Houthooft, R., Schulman, J., Sutskever, I., Abbeel, P.: InfoGAN: interpretable representation learning by information maximizing generative adversarial nets. Adv. Neural Inf. Process. Syst. 29 (2016)
Chen, Z., et al.: Semantics disentangling for generalized zero-shot learning. In: Proceedings of the IEEE/CVF International Conference on Computer Vision, pp. 8712–8720 (2021)
Di, H., Ke, X., Peng, Z., Dongdong, Z.: Surface defect classification of steels with a new semi-supervised learning method. Opt. Lasers Eng. 117, 40–48 (2019)
Goodfellow, I., et al.: Generative adversarial nets. Adv. Neural Inf. Process. Syst. 27 (2014)
Gulrajani, I., Ahmed, F., Arjovsky, M., Dumoulin, V., Courville, A.C.: Improved training of Wasserstein GANs. Adv. Neural Inf. Process. Syst. 30 (2017)
He, K., Zhang, X., Ren, S., Sun, J.: Deep residual learning for image recognition. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 770–778 (2016)
Huang, Y., Qiu, C., Yuan, K.: Surface defect saliency of magnetic tile. Vis. Comput. 36(1), 85–96 (2018). https://doi.org/10.1007/s00371-018-1588-5
Kingma, D.P., Ba, J.: Adam: a method for stochastic optimization. arXiv preprint arXiv:1412.6980 (2014)
Koksal, A., Lu, S.: RF-GAN: a light and reconfigurable network for unpaired image-to-image translation. In: Proceedings of the Asian Conference on Computer Vision (2020)
LeCun, Y., Bengio, Y., Hinton, G.: Deep learning. Nature 521(7553), 436–444 (2015)
Liu, J., Wang, C., Su, H., Du, B., Tao, D.: Multistage GAN for fabric defect detection. IEEE Trans. Image Process. 29, 3388–3400 (2019)
Liu, L., Cao, D., Wu, Y., Wei, T.: Defective samples simulation through adversarial training for automatic surface inspection. Neurocomputing 360, 230–245 (2019)
Mao, X., Li, Q., Xie, H., Lau, R.Y., Wang, Z., Paul Smolley, S.: Least squares generative adversarial networks. In: Proceedings of the IEEE International Conference on Computer Vision, pp. 2794–2802 (2017)
Naranjo-Alcazar, J., Perez-Castanos, S., Zuccarello, P., Cobos, M.: Acoustic scene classification with squeeze-excitation residual networks. IEEE Access 8, 112287–112296 (2020)
Niu, S., Li, B., Wang, X., Lin, H.: Defect image sample generation with GAN for improving defect recognition. IEEE Trans. Autom. Sci. Eng. 17(3), 1611–1622 (2020)
Park, J.-K., Kwon, B.-K., Park, J.-H., Kang, D.-J.: Machine learning-based imaging system for surface defect inspection. Int. J. Precis. Eng. Manuf.-Green Technol. 3(3), 303–310 (2016). https://doi.org/10.1007/s40684-016-0039-x
Radford, A., Metz, L., Chintala, S.: Unsupervised representation learning with deep convolutional generative adversarial networks. arXiv preprint arXiv:1511.06434 (2015)
Schroff, F., Kalenichenko, D., Philbin, J.: FaceNet: a unified embedding for face recognition and clustering. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 815–823 (2015)
Simonyan, K., Zisserman, A.: Very deep convolutional networks for large-scale image recognition. arXiv preprint arXiv:1409.1556 (2014)
Xian, Y., Sharma, S., Schiele, B., Akata, Z.: F-VAEGAN-D2: a feature generating framework for any-shot learning. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 10275–10284 (2019)
Xuan, Q., Chen, Z., Liu, Y., Huang, H., Bao, G., Zhang, D.: Multiview generative adversarial network and its application in pearl classification. IEEE Trans. Industr. Electron. 66(10), 8244–8252 (2018)
Zhang, G., Cui, K., Hung, T.Y., Lu, S.: Defect-GAN: high-fidelity defect synthesis for automated defect inspection. In: Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, pp. 2524–2534 (2021)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2022 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Chen, Y., Hu, R., Wang, Z., Li, Y. (2022). How to Face Unseen Defects? UDGAN for Improving Unseen Defects Recognition. In: Pimenidis, E., Angelov, P., Jayne, C., Papaleonidas, A., Aydin, M. (eds) Artificial Neural Networks and Machine Learning – ICANN 2022. ICANN 2022. Lecture Notes in Computer Science, vol 13529. Springer, Cham. https://doi.org/10.1007/978-3-031-15919-0_16
Download citation
DOI: https://doi.org/10.1007/978-3-031-15919-0_16
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-15918-3
Online ISBN: 978-3-031-15919-0
eBook Packages: Computer ScienceComputer Science (R0)