Mechanisms of Generative Image-to-Image Translation Networks

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Chen, Guangzong, Sun, Mingui, Mao, Zhi-Hong, Liu, Kangni, Jia, Wenyan
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917858663989248
author Chen, Guangzong
Sun, Mingui
Mao, Zhi-Hong
Liu, Kangni
Jia, Wenyan
author_facet Chen, Guangzong
Sun, Mingui
Mao, Zhi-Hong
Liu, Kangni
Jia, Wenyan
contents Generative Adversarial Networks (GANs) are a class of neural networks that have been widely used in the field of image-to-image translation. In this paper, we propose a streamlined image-to-image translation network with a simpler architecture compared to existing models. We investigate the relationship between GANs and autoencoders and provide an explanation for the efficacy of employing only the GAN component for tasks involving image translation. We show that adversarial for GAN models yields results comparable to those of existing methods without additional complex loss penalties. Subsequently, we elucidate the rationale behind this phenomenon. We also incorporate experimental results to demonstrate the validity of our findings.
format Preprint
id arxiv_https___arxiv_org_abs_2411_10368
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Mechanisms of Generative Image-to-Image Translation Networks
Chen, Guangzong
Sun, Mingui
Mao, Zhi-Hong
Liu, Kangni
Jia, Wenyan
Computer Vision and Pattern Recognition
Artificial Intelligence
Image and Video Processing
Generative Adversarial Networks (GANs) are a class of neural networks that have been widely used in the field of image-to-image translation. In this paper, we propose a streamlined image-to-image translation network with a simpler architecture compared to existing models. We investigate the relationship between GANs and autoencoders and provide an explanation for the efficacy of employing only the GAN component for tasks involving image translation. We show that adversarial for GAN models yields results comparable to those of existing methods without additional complex loss penalties. Subsequently, we elucidate the rationale behind this phenomenon. We also incorporate experimental results to demonstrate the validity of our findings.
title Mechanisms of Generative Image-to-Image Translation Networks
topic Computer Vision and Pattern Recognition
Artificial Intelligence
Image and Video Processing
url https://arxiv.org/abs/2411.10368