A Structure-Aware Deep Learning Network for the Transfer of Chinese Landscape Painting Style

Der Lor Way*, Chang Hao Lo, Yu Hsien Wei, Zen Chung Shih

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Recently, deep learning technology has made a breakthrough in computer vision, image processing, and other fields. Some researchers suggested neural style transfer method using a convolutional neural network (CNN). They established the correlation of features in a neural network to be treated as the style. However, their performance is unacceptable for Chinese landscape painting. According to the property of the Chinese landscape painting, this paper proposes a novel two stage style transfer method that imitates multiple styles of Chinese landscape painting based on deep learning. The structure of an input photo was simplified in the first stage. Then, the result of the first stage was transferred into the final stylized image in second stage. A generative adversarial network (GAN) is applied to train in each stage. Besides, a novel loss function was proposed to keep the shape of the content image. Finally, our method haves successfully imitated several styles of Chinese Landscape ink painting.

Original languageEnglish
Title of host publicationCulture and Computing - 11th International Conference, C and C 2023, Held as Part of the 25th HCI International Conference, HCII 2023, Proceedings
EditorsMatthias Rauterberg
PublisherSpringer Science and Business Media Deutschland GmbH
Pages326-337
Number of pages12
ISBN (Print)9783031347313
DOIs
StatePublished - 2023
Event11th International Conference on Culture and Computing, C and C 2023, held as part of the 25th International Conference on Human-Computer Interaction, HCII 2023 - Copenhagen, Denmark
Duration: 23 Jul 202328 Jul 2023

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume14035 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference11th International Conference on Culture and Computing, C and C 2023, held as part of the 25th International Conference on Human-Computer Interaction, HCII 2023
Country/TerritoryDenmark
CityCopenhagen
Period23/07/2328/07/23

Keywords

  • Chinese landscape painting
  • generative adversarial network (GAN)
  • loss function
  • style transfer

Fingerprint

Dive into the research topics of 'A Structure-Aware Deep Learning Network for the Transfer of Chinese Landscape Painting Style'. Together they form a unique fingerprint.

Cite this