基于对抗生成网络的纹理合成方法
Texture synthesis method based on generative adversarial networks
查看参考文献15篇
文摘
|
纹理合成是计算机图形学、计算机视觉和图像处理领域的研究热点之一。传统的纹理合成方法往往通过提取有效的特征样式或统计量并在该特征信息的约束下生成随机图像来实现。对抗生成网络作为一种较新的深度网络形式,通过生成器和判别器的对抗训练能够随机生成与观测数据具有相同分布的新数据。鉴于此,提出了一种基于对抗生成网络的纹理合成方法。该算法的优点是不需要经过多次迭代就能够生成更真实纹理图像,且生成图像在视觉上与观测纹理图像一致的同时具有一定随机性。一系列针对随机纹理和结构性纹理的合成实验验证了该算法的有效性。 |
其他语种文摘
|
Texture synthesis is a hot research topic in the fields of computer graphics, vision, and image processing. Traditional texture synthesis methods are generally achieved by extracting effective feature patterns or statistics and generating random images under the constraint of the feature information. Generative adversarial networks (GANs) is a new type of deep network. It can randomly generate new data of the same distribution as the observed data by training generator and discriminator in an adversarial learning mechanism. Inspired by this point, a texture synthesis method based on GANs was proposed. The advantage of the algorithm was that it could generate more realistic texture images without iteration; the generated images were visually consistent with the observed texture image and also had randomness. A series of experiments for random texture and structured texture synthesis verify the effectiveness of the proposed algorithm. |
来源
|
红外与激光工程
,2018,47(2):0203005-1-0203005-6 【核心库】
|
DOI
|
10.3788/IRLA201847.0203005
|
关键词
|
纹理合成
;
深度学习
;
生成模型
;
对抗生成网络
|
地址
|
1.
东北大学信息科学与工程学院, 辽宁, 沈阳, 110000
2.
中国科学院沈阳自动化研究所, 机器人学国家重点实验室, 辽宁, 沈阳, 110000
|
语种
|
中文 |
文献类型
|
研究性论文 |
ISSN
|
1007-2276 |
学科
|
自动化技术、计算机技术 |
基金
|
国家自然科学基金
;
中国科学院青年创新促进会项目
|
文献收藏号
|
CSCD:6207497
|
参考文献 共
15
共1页
|
1.
Huang Jun. Surfaces texture synthesis based on texel distribution.
Journal of Chinese Computer Systems. (in Chinese),2016,37(10):2361-2365
|
被引
1
次
|
|
|
|
2.
Criminisi A. Region filling and object removal by exemplar-based image inpainting.
IEEE Tras Process,2004,13(9):1200-1212
|
被引
358
次
|
|
|
|
3.
Kwatra V. Graphcut textures: Image and video synthesis using graph cuts.
ACM Transactions on Graphics,2003:277-286
|
被引
100
次
|
|
|
|
4.
Lefebvre S. Parallel controllable texture synthesis.
ACM Transactions on Graphics,2005,24(3):777-786
|
被引
26
次
|
|
|
|
5.
Zhang Weiwei. Texture synthesis method by adaptive selecting size of patches.
Computer Engineering and Applications. (in Chinese),2012,48(17):170-173
|
被引
1
次
|
|
|
|
6.
Song C Z. Filters, random fields and maximum entropy (FRAME): towards a unified theory for texture modeling.
International Journal of Computer Vision,1998,27(2):107-126
|
被引
1
次
|
|
|
|
7.
Kwatra V. Texture optimization for example-based synthesis.
ACM Transactions on Graphics,2005:795-802
|
被引
58
次
|
|
|
|
8.
Urs R D. Maximum-likelihood based synthesis of volumetric textures from a 2D sample.
IEEE Transactions on Image Processing,2014,23(4):1820-1830
|
被引
2
次
|
|
|
|
9.
Xie J. Learning sparse FRAME models for natural image patterns.
International Journal of Computer Vision,2015,114(2/3):1-22
|
被引
1
次
|
|
|
|
10.
Lu Y. Learning FRAME models using CNN filters.
Computer Science,2015
|
被引
1
次
|
|
|
|
11.
Gatys L A. Texture synthesis using convolutional neural networks.
Febs Letters,2015,70(1):51-55
|
被引
4
次
|
|
|
|
12.
Schreiber S. Texture synthesis using convolutional neural networks with long-range consistency and spectral constraints.
Pattern Recognition Association of South Africa and Robotics and Mechatronics International Conference,2017:1-6
|
被引
1
次
|
|
|
|
13.
Xiao Chunxia. Global texture optimization incorporating with image detail.
Chinese Journal of Computers. (in Chinese),2009,32(6):1196-1205
|
被引
1
次
|
|
|
|
14.
Tang Ying. GPU-based texture synthesis with preserved structures.
Computer Science. (in Chinese),2016,43(4):299-302
|
被引
1
次
|
|
|
|
15.
Goodfellow I J. Generative adversarial nets.
International Conference on Neural Information Processing Systems,2014:2672-2680
|
被引
56
次
|
|
|
|
|