Browse > Article
http://dx.doi.org/10.17661/jkiiect.2022.15.5.380

A Study on the Image Preprosessing model linkage method for usability of Pix2Pix  

Kim, Hyo-Kwan (Department of Fintech Korea Polytechnics)
Hwang, Won-Yong (Department of Fintech Korea Polytechnics)
Publication Information
The Journal of Korea Institute of Information, Electronics, and Communication Technology / v.15, no.5, 2022 , pp. 380-386 More about this Journal
Abstract
This paper proposes a method for structuring the preprocessing process of a training image when color is applied using Pix2Pix, one of the adversarial generative neural network techniques. This paper concentrate on the prediction result can be damaged according to the degree of light reflection of the training image. Therefore, image preprocesisng and parameters for model optimization were configured before model application. In order to increase the image resolution of training and prediction results, it is necessary to modify the of the model so this part is designed to be tuned with parameters. In addition, in this paper, the logic that processes only the part where the prediction result is damaged by light reflection is configured together, and the pre-processing logic that does not distort the prediction result is also configured.Therefore, in order to improve the usability, the accuracy was improved through experiments on the part that applies the light reflection tuning filter to the training image of the Pix2Pix model and the parameter configuration.
Keywords
Discriminator; Generator; GAN; Pix2Pix; Homomorphic Filter;
Citations & Related Records
연도 인용수 순위
  • Reference
1 P. Isola et al., "Image-to-Image Translation with Conditional Adversarial Nets", IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 1127-1128, 2017
2 Ronneberger O, U-net: Convolutional net-works for biomedical image segmentation, Int'l Conf. on Medi-cal Image Computing and Computer-assisted Intervention, pp. 235-237, 2015
3 Ronneberger O, U-net: Convolutional net-works for biomedical image segmentation,Int'l Conf. on Medi-cal Image Computing and Computer-assisted Intervention, 2015
4 Ugur Demir,Gozde Unal, Patch-Based Image Inpainting with Generative Adversarial Networks,arXiv:1803.07422v1, pp1-3, 2018
5 MARAGOS, Petros, Morphological filtering. In: The essential guide to image processing, Academic Press, pp293-321, 2009
6 Yanyun Qu, Yizi Chen, Jingying Huang, Yuan Xie, "Enhanced Pix2pix Dehazing Network", Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp. 8160-8168, 2019
7 Milletari F, V-net: Fully convolutional neural networks for volumetric medical image segmentation, IEEE Int'l Conf. on 3D vision (3DV), pp. 565-565, 2016
8 Phillip Isola, https://github.com/phillipi/pix2pix/blob/master/scripts/receptive_field_sizes.m, 2014
9 Google Tensorflow, https://www.tensorflow.org/tutorials/generative/pix2pix?hl=ko, 2022