GAN-based data augmentation of prohibited item X-ray images in security inspection

Abstract

Convolutional neural networks (CNNs) based methods for automatic discriminant of prohibited items in X-ray images attract attention increasingly. However, it is difficult to train a reliable CNN model using the available X-ray security image databases, since they are not enough in sample quantity and diversity. Recently, generative adversarial network (GAN) has been widely used in image generation and regarded as a power model for data augmentation. In this paper, we propose a data augmentation method for X-ray prohibited item images based on GAN. First, the network structure and loss function of the self-attention generative adversarial network (SAGAN) are improved to generate the realistic X-ray prohibited item images. Then, the images generated by our model are evaluated using GAN-train and GAN-test. Experimental results of GAN-train and GAN-test are 99.91% and 98.82% respectively. It implies that our model can enlarge the X-ray prohibited item image database effectively.

References

[1] AkcayS, KundegorskiM E, DevereuxM, BreckonY P. Transfer Learning Using Convolutional Neural Networks for Object Classification within X-Ray Baggage Security Imagery, 2016,

[2] GirshickR. Fast R-CNN, IEEE International Conference on Computer Vision, 2015, 1440

[3] AkcayS, KundegorskiM E, WillcocksC G, BreckonY P. IEEE Transactions on Information Forensics and Security, 2018, 13: 2203

[4] XuM-s, ZhangH-g, YangJ-f. Prohibited Item Detection in Airport X-Ray Security Images via Attention Mechanism Based CNN, 2018, 429

[5] MeryD, SvecE, AriasM, RiffoV, SaavedraJ M, BanerjeeS. IEEE Transactions on Systems, Man, and Cybernetics, 2017, 47: 682

[6] MiaoC, XieL, WanF, SuC, LiuH, JiaoJ a, YeQ. SIXray: A Large-scale Security Inspection X-ray Benchmark for Prohibited Item Discovery, Computer Vision and Pattern Recognition, 2019, 2119

[7] GoodfellowI, Pouget-AbadieJ, MirzaM, XuB, Warde-FarleyD, OzairS, CourvilleA, BengioY. Generative Adversarial Nets, the 27th International Conference on Neural Information Processing Systems, 2014, 2672

[8] RadfordA, MetzL, ChintalaS. Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks, 2015,

[9] GulrajaniI, AhmedF, ArjovskyM, DumoulinV, CourvilleA C. Improved Training of Wasserstein GANs, 2017,

[10] ZhangH, GoodfellowI, MetaxasD, OdenaA S. elf-Attention Generative Adversarial Networks, 2018,

[11] BrockA, DonahueJ, SimonyanK. Large Scale GAN Training for High Fidelity Natural Image Synthesis, 2018,

[12] AntoniouA, AmosS, HarrisonE. Data Augmentation Generative Adversarial Networks, 2017,

[13] Frid-AdarM, KlangE, AmitaiM, GoldbergerJ, GreenspanH. Synthetic Data Augmentation Using GAN for Improved Liver Lesion Classification, 15th International Symposium on Biomedical Imaging, 2018, 289

[14] ZhaoZ H, ZhangH G, YangJ F. A GAN-Based Image Generation Method for X-Ray Security Prohibited Items, 2018, 420

[15] HeuseH, RamsauerH, UnterthinerT, NesslerB, HochreiterS. GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash Equilibrium, 2017, 6629

[16] ShmelkovK, SchmidC, AlahariK. How Good is My GAN? European Conference on Computer Vision, 2018, 213

[17] WangX, GirshickR, GuptaA, HeK-m. Non-local Neural Networks, IEEE Conference on Computer Vision and Pattern Recognition, 2018, 7794

[18] LimJ H, YeJ C. Geometric GAN, 2017,

[19] SalimansT, GoodfellowI, ZarembaW, CheungV, RadfordA, ChenX. Improved Techniques for Training GANs, Neural Information Processing Systems, 2016, 2234