Multi-Stage Generation of Tile Images Based on Generative Adversarial Network
Deep learning techniques have been recently widely used in the field of texture image generation. There are still two major problems when applying them to tile image design work. On the one hand, there is still lack of enough diverse ceramic tile images for the training process. On the other hand, t...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2022-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/9933808/ |
| _version_ | 1811199303344979968 |
|---|---|
| author | Jianfeng Lu Mengtao Shi Yuhang Lu Ching-Chun Chang Li Li Rui Bai |
| author_facet | Jianfeng Lu Mengtao Shi Yuhang Lu Ching-Chun Chang Li Li Rui Bai |
| author_sort | Jianfeng Lu |
| collection | DOAJ |
| description | Deep learning techniques have been recently widely used in the field of texture image generation. There are still two major problems when applying them to tile image design work. On the one hand, there is still lack of enough diverse ceramic tile images for the training process. On the other hand, the output image is difficult to control and adjust, and cannot meet the designer’s requirements of interactivity. Therefore, we propose a multi-stage generation algorithm of tile images based on generative adversarial network(GAN). First, the multi-scale attention GAN is applied to generate controllable texture image. Then, the SWAG texture synthesis GAN is also applied to obtain controllable and diverse image style. And finally, through the style iteration mechanism and the multiple step magnification method based on image super-resolution reconstruction network, the final tile images can be automatically generated with larger-size and higher-precision. The relevant experiments demonstrate that our method can not only generate high-quality tile images in a relatively short period of time, but also consider human interaction to a certain extent, and maintain a certain degree of control over the main texture and style of the final generated tile images. It has good and wide application value. |
| first_indexed | 2024-04-12T01:45:43Z |
| format | Article |
| id | doaj.art-13a84df5f6144f24b58783b0473ee04a |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-04-12T01:45:43Z |
| publishDate | 2022-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-13a84df5f6144f24b58783b0473ee04a2022-12-22T03:53:05ZengIEEEIEEE Access2169-35362022-01-011012750212751310.1109/ACCESS.2022.32186369933808Multi-Stage Generation of Tile Images Based on Generative Adversarial NetworkJianfeng Lu0https://orcid.org/0000-0002-3636-6528Mengtao Shi1https://orcid.org/0000-0002-7327-1034Yuhang Lu2https://orcid.org/0000-0002-2408-5825Ching-Chun Chang3https://orcid.org/0000-0001-7723-4591Li Li4https://orcid.org/0000-0002-5453-226XRui Bai5https://orcid.org/0000-0003-4773-8101College of Computer Science, Hangzhou Dianzi University, Hangzhou, ChinaCollege of Computer Science, Hangzhou Dianzi University, Hangzhou, ChinaCollege of Computer Science, Hangzhou Dianzi University, Hangzhou, ChinaDepartment of Computer Science, University of Warwick, Coventry, U.KCollege of Computer Science, Hangzhou Dianzi University, Hangzhou, ChinaKey Laboratory of Brain Machine Collaborative Intelligence of Zhejiang Province, Hangzhou, ChinaDeep learning techniques have been recently widely used in the field of texture image generation. There are still two major problems when applying them to tile image design work. On the one hand, there is still lack of enough diverse ceramic tile images for the training process. On the other hand, the output image is difficult to control and adjust, and cannot meet the designer’s requirements of interactivity. Therefore, we propose a multi-stage generation algorithm of tile images based on generative adversarial network(GAN). First, the multi-scale attention GAN is applied to generate controllable texture image. Then, the SWAG texture synthesis GAN is also applied to obtain controllable and diverse image style. And finally, through the style iteration mechanism and the multiple step magnification method based on image super-resolution reconstruction network, the final tile images can be automatically generated with larger-size and higher-precision. The relevant experiments demonstrate that our method can not only generate high-quality tile images in a relatively short period of time, but also consider human interaction to a certain extent, and maintain a certain degree of control over the main texture and style of the final generated tile images. It has good and wide application value.https://ieeexplore.ieee.org/document/9933808/Tile imagesgenerative adversarial networksstyle transferimage super-resolution magnification |
| spellingShingle | Jianfeng Lu Mengtao Shi Yuhang Lu Ching-Chun Chang Li Li Rui Bai Multi-Stage Generation of Tile Images Based on Generative Adversarial Network IEEE Access Tile images generative adversarial networks style transfer image super-resolution magnification |
| title | Multi-Stage Generation of Tile Images Based on Generative Adversarial Network |
| title_full | Multi-Stage Generation of Tile Images Based on Generative Adversarial Network |
| title_fullStr | Multi-Stage Generation of Tile Images Based on Generative Adversarial Network |
| title_full_unstemmed | Multi-Stage Generation of Tile Images Based on Generative Adversarial Network |
| title_short | Multi-Stage Generation of Tile Images Based on Generative Adversarial Network |
| title_sort | multi stage generation of tile images based on generative adversarial network |
| topic | Tile images generative adversarial networks style transfer image super-resolution magnification |
| url | https://ieeexplore.ieee.org/document/9933808/ |
| work_keys_str_mv | AT jianfenglu multistagegenerationoftileimagesbasedongenerativeadversarialnetwork AT mengtaoshi multistagegenerationoftileimagesbasedongenerativeadversarialnetwork AT yuhanglu multistagegenerationoftileimagesbasedongenerativeadversarialnetwork AT chingchunchang multistagegenerationoftileimagesbasedongenerativeadversarialnetwork AT lili multistagegenerationoftileimagesbasedongenerativeadversarialnetwork AT ruibai multistagegenerationoftileimagesbasedongenerativeadversarialnetwork |