Image to image translation(I2I) is one of the important parts in computer vision area. It includes lots of applications in this time. GAN provides basic theory for this problem which is an advanced method of machine learning area. The structure of GAN contains two main parts: Generator and Discriminator. This new method developed the performance of I2I problem. In this paper, five papers of I2I area using GAN method have been summarized including a paper about cGAN and a paper about cycle GAN. The rest three paper is about unsupervised learning. These five papers used different method based on GAN algorithm and can be used on different problems. Finally, there are still some problems cannot be solved in this area, this paper also discuss which problem could be development in the future.
Comparison on Image to Image Translation Algorithms
2022-10-12
1712669 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Comparison of image reconstruction algorithms
SPIE | 1995
|Unsupervised Image Translation
British Library Conference Proceedings | 2003
|Unsupervised image translation
IEEE | 2003
|Image translation to enhance IR2VIS image registration
SPIE | 2021
|