Style Transfer: From Stitching to Neural Networks

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Xu, Xinhe, Wang, Zhuoer, Zhang, Yihan, Liu, Yizhou, Wang, Zhaoyue, Xu, Zhihao, Zhao, Muhan, Luo, Huaiying
Formato: Preprint
Publicado: 2024
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866910694952140800
author Xu, Xinhe
Wang, Zhuoer
Zhang, Yihan
Liu, Yizhou
Wang, Zhaoyue
Xu, Zhihao
Zhao, Muhan
Luo, Huaiying
author_facet Xu, Xinhe
Wang, Zhuoer
Zhang, Yihan
Liu, Yizhou
Wang, Zhaoyue
Xu, Zhihao
Zhao, Muhan
Luo, Huaiying
contents This article compares two style transfer methods in image processing: the traditional method, which synthesizes new images by stitching together small patches from existing images, and a modern machine learning-based approach that uses a segmentation network to isolate foreground objects and apply style transfer solely to the background. The traditional method excels in creating artistic abstractions but can struggle with seamlessness, whereas the machine learning method preserves the integrity of foreground elements while enhancing the background, offering improved aesthetic quality and computational efficiency. Our study indicates that machine learning-based methods are more suited for real-world applications where detail preservation in foreground elements is essential.
format Preprint
id arxiv_https___arxiv_org_abs_2409_00606
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Style Transfer: From Stitching to Neural Networks
Xu, Xinhe
Wang, Zhuoer
Zhang, Yihan
Liu, Yizhou
Wang, Zhaoyue
Xu, Zhihao
Zhao, Muhan
Luo, Huaiying
Computer Vision and Pattern Recognition
This article compares two style transfer methods in image processing: the traditional method, which synthesizes new images by stitching together small patches from existing images, and a modern machine learning-based approach that uses a segmentation network to isolate foreground objects and apply style transfer solely to the background. The traditional method excels in creating artistic abstractions but can struggle with seamlessness, whereas the machine learning method preserves the integrity of foreground elements while enhancing the background, offering improved aesthetic quality and computational efficiency. Our study indicates that machine learning-based methods are more suited for real-world applications where detail preservation in foreground elements is essential.
title Style Transfer: From Stitching to Neural Networks
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2409.00606