Cycle Generative Adversarial Networks Algorithm With Style Transfer For Image Generation

The biggest challenge faced by a Machine Learning Engineer is the lack of data they have, especially for 2-dimensional images. The image is processed to be trained into a Machine Learning model so that it can recognize patterns in the data and provide predictions. This research is intended to create...

Full description

Saved in:
Bibliographic Details
Published inarXiv.org
Main Authors Anugrah Akbar Praramadhan, Guntur Eka Saputra
Format Paper
LanguageEnglish
Published Ithaca Cornell University Library, arXiv.org 11.01.2021
Subjects
Algorithms
Astronomy
Generative adversarial networks
Image processing
Machine learning
Pattern recognition
Online AccessGet full text

Cover

More Information
Summary:The biggest challenge faced by a Machine Learning Engineer is the lack of data they have, especially for 2-dimensional images. The image is processed to be trained into a Machine Learning model so that it can recognize patterns in the data and provide predictions. This research is intended to create a solution using the Cycle Generative Adversarial Networks (GANs) algorithm in overcoming the problem of lack of data. Then use Style Transfer to be able to generate a new image based on the given style. Based on the results of testing the resulting model has been carried out several improvements, previously the loss value of the photo generator: 3.1267, monet style generator: 3.2026, photo discriminator: 0.6325, and monet style discriminator: 0.6931 to photo generator: 2.3792, monet style generator: 2.7291, photo discriminator: 0.5956, and monet style discriminator: 0.4940. It is hoped that the research will make the application of this solution useful in the fields of Education, Arts, Information Technology, Medicine, Astronomy, Automotive and other important fields.
ISSN:2331-8422