Data augmentation method for light guide plate based on improved CycleGAN

Abstract

An improved cycle-consistent generative adversarial network (CycleGAN) method for defect data augmentation based on feature fusion and self attention residual module is proposed to address the insufficiency of defect sample data for light guide plate (LGP) in production, as well as the problem of minor defects. Two optimizations are made to the generator of CycleGAN: fusion of low resolution features obtained from partial up-sampling and down-sampling with high-resolution features, combination of self attention mechanism with residual network structure to replace the original residual module. Qualitative and quantitative experiments were conducted to compare different data augmentation methods, and the results show that the defect images of the LGP generated by the improved network were more realistic, and the accuracy of the you only look once version 5 (YOLOv5) detection network for the LGP was improved by 5.6%, proving the effectiveness and accuracy of the proposed method.

Keywords

A

References

[1] JavaidM, HaleemA, SinghR P, et al.. Exploring impact and features of machine vision for progressive industry 4.0 culture. Sensors international, 2022, 3: 100-132[J]

[2] RADFORD A, METZ L, CHINTALA S. Unsupervised representation learning with deep convolutional generative adversarial networks[EB/OL]. (2015-11-19) [2024-01-23]. http://arxiv.org/abs/1511.06434?context=cs.LG.htm.

[3] LinZ P, ZengL B, WuQ S. Cervical cell image data enhancement based on generative adversarial network. Science technology and engineering, 2020, 20(28): 11672-11677[J]

[4] WangJ N, SuJ, YangK. Insulator image generation method based on Cycle-GAN. Guangdong electric power, 2020, 33(01): 100-108[J]

[5] ZhuJ Y, ParkT, IsolaP, et al.. Unpaired image-to-image translation using cycle-consistent adversarial networks. Proceedings of the IEEE International Conference on Computer Vision, October 22–29, 2017, Venice, Italy, 2017, New York. IEEE. 22232232[C]

[6] SONG Z W, YAO H, TIAN D, et al. Improved Cycle-GAN for super-resolution of engineering drawings[J]. Measurement science and technology, 2023, 34(7).

[7] ChoiY, ChoiM, KimM, et al.. StarGAN: unified generative adversarial networks for multi-domain image-to-image translation. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, June 18–23, 2018, Salt Lake City, UT, USA, 2018, New York. IEEE. 87898797[C]

[8] ThambawitaV, SalehiP, SheshkalS A, et al.. SinGAN-Seg: synthetic training data generation for medical image segmentation. PloS one, 2022, 175e0267976[J]

[9] ShenY, HuangR, HuangW. GD-StarGAN: multi-domain image-to-image translation in garment design. PloS one, 2020, 154e0231719[J]

[10] LuceyP, CohnJ F, KanadeT, et al.. The extended cohn-kanade dataset (CK+): a complete dataset for action unit and emotion-specified expression. IEEE Computer Society Conference on Computer Vision and Pattern Recognition-Workshops, June 13–18, 2010, San Francisco, CA, USA, 2010, New York. IEEE. 94101[C]

[11] SunX, DingX L. Facial expression data enhancement method based on generative adversarial network. Computer engineering and applications, 2020, 56(04): 115-121[J]

[12] LiuK, WenX, HuangM, et al.. Solar cell defect enhancement method based on generative adversarial network. Journal of Zhejiang University (engineering science), 2020, 54(4): 684-693[J]

[13] YANG Z Z, SHAO J, YANG Y P. An improved CycleGAN for data augmentation in person re-identification[J]. Big data research, 2023, 34(28).

[14] ZhuL, WangX, KeZ, et al.. BiFormer: vision transformer with bi-level routing attention. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, June 18–22, 2023, Vancouver, Canada, 2023, New York. IEEE. 1032310333[C]

[15] GLENN J. Ultralytics/yolov5 v6.0[EB/OL]. (2021-10-21) [2024-01-23]. https://github.com/ultralytics/yolov5/releases/tag/v6.0.

RIGHTS & PERMISSIONS

Tianjin University of Technology