CycleGAN and SRGAN to Enrich the Dataset
DOI:
10.33395/sinkron.v7i2.11384Keywords:
Cycle GAN, Super Resolution GAN, Deep Learning, Data acquisition, Varian GANAbstract
When developments in the field of computer science are growing rapidly. For example, the development of image or video predictions for various fields has been widely applied to assist further processes. The field of computer vision has created many ideas about processing using deep learning algorithms. Sometimes the problem with using deep learning or machine learning is in the availability of the dataset or the unavailability of the dataset. Various methods are used to add to or enrich the dataset. One way is to add an image dataset by creating a synthetic image. One of the well-known algorithms is Generative Adversarial Networks as an algorithm for generating synthetic images. Currently, there are many variations of the GAN to around 500 variants. This research is to utilize the Cycle GAN architecture in order to enrich the dataset. By doing GAN as a synthetic image generator. This is very important in procuring image datasets, for training and testing models of Deep Learning algorithms such as Convolutional Neural Networks. In addition, the use of synthetic images produces a deep learning model to avoid overfitting. One of the causes of the overfitting problem is the lack of datasets. There are many ways to add image datasets, by cropping, continuously rotating 90 degrees, 180 degrees. The reason for using Cycle Generative Adversarial Networks is because this method is not as complicated as other GANs, but also not as simple. Cycle GAN synthetic images are processed with Super Resolution GAN, which aims to clarify image quality. So that it produces a different image and good image quality.
Downloads
References
Akbar, A., Praramadhan, G., & Eka, S. (2021). Cycle Generative Adversarial Networks Algorithm With Style Transfer For Image Generation. Computer Vision and Pattern Recognition. https://doi.org/https://doi.org/10.48550/arXiv.2101.03921
Asaf, B.-E., Dana, C., Noa, C., & Hayit, G. (2021). Improved CycleGAN with application to COVID-19 classification. Medical Imaging. https://doi.org/https://doi.org/10.1117/12.2582162
Barth, R., Hemming, J., & Van Henten, E. J. (2020). Optimising realism of synthetic images using cycle generative adversarial networks for improved part segmentation. Computers and Electronics in Agriculture, 173(March), 105378. https://doi.org/10.1016/j.compag.2020.105378
Efros, J.-Y., Zhu;, Taesung, P., Phillip, I., & A, A. (2017). Unpaired Image-to-Image Translationusing Cycle-Consistent Adversarial Networks. Computer Vision and Pattern Recognition. https://doi.org/https://doi.org/10.48550/arXiv.1703.10593
Goodfellow, I., Mehdi, M., Bing, X., & David, W. (2014). Generative Adversarial Nets.
Lan, L., You, L., Zhang, Z., Fan, Z., Zhao, W., Zeng, N., Chen, Y., & Zhou, X. (2020). Generative Adversarial Networks and Its Applications in Biomedical Informatics. Frontiers in Public Health, 8(May), 1–14. https://doi.org/10.3389/fpubh.2020.00164
Ledig, C., Lucas, T., Ferenc, H. ́ar;, Jose, C., Andrew, C., Alejandro, A., Andrew, A., Alykhan, T., Johannes, T., Zehan, W., & Wenzhe, S. (2017). Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network.
Lee, Y., Yun, J., Hong, Y., Lee, J., & Jeon, M. (2018). Accurate License Plate Recognition and Super-Resolution Using a Generative Adversarial Networks on Traffic Surveillance Video. 2018 IEEE International Conference on Consumer Electronics - Asia, ICCE-Asia 2018, June 2018, 1–4. https://doi.org/10.1109/ICCE-ASIA.2018.8552121
Maqsood, M. H., Mumtaz, R., Haq, I. U., Shafi, U., Zaidi, S. M. H., & Hafeez, M. (2021). Super resolution generative adversarial network (Srgans) for wheat stripe rust classification. Sensors, 21(23), 1–12. https://doi.org/10.3390/s21237903
Nasrollahi, K., & Moeslund, T. B. (2014). Super-resolution: A comprehensive survey. In Machine Vision and Applications (Vol. 25, Issue 6). https://doi.org/10.1007/s00138-014-0623-4
Park, J., Han, D. K., & Ko, H. (2019). Adaptive weighted multi-discriminator CycleGAN for underwater image enhancement. In Journal of Marine Science and Engineering (Vol. 7, Issue 7). https://doi.org/10.3390/jmse7070200
Suarez, P. L., Sappa, A. D., Vintimilla, B. X., & Hammoud, R. I. (2019). Image vegetation index through a cycle generative adversarial network. IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, 2019-June(1), 1014–1021. https://doi.org/10.1109/CVPRW.2019.00133
Tang, C., Li, J., Wang, L., Li, Z., Jiang, L., Cai, A., Zhang, W., Liang, N., Li, L., & Yan, B. (2019). Unpaired Low-Dose CT Denoising Network Based on Cycle-Consistent Generative Adversarial Network with Prior Image Information. Computational and Mathematical Methods in Medicine, 2019. https://doi.org/10.1155/2019/8639825
Wenlong, Z., Yihao, L., Chao, D., & Yu, Q. (2021). RankSRGAN- Super Resolution GenerativeAdversarial Networks with Learning to Rank. JOURNAL OF LATEX CLASS FILES.
Won, T. (2022). An Experiment on Image Restoration Applying the Cycle Generative Adversarial Network to Partial Occlusion Kompsat-3A Image. 38(1), 33–43.
Yang, Q., Yang, R., Davis, J., & Nist, D. (n.d.). Spatial-Depth Super Resolution for Range Images.
Downloads
How to Cite
Issue
Section
License
Copyright (c) 2022 Budi Priswanto, Handri Santoso

This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.