[1]周传华,吴幸运,李 鸣.基于 WGAN 单帧人脸图像超分辨率算法[J].计算机技术与发展,2020,30(09):29-35.[doi:10. 3969 / j. issn. 1673-629X. 2020. 09. 006]
　ZHOU Chuan-hua,WU Xing-yun,LI Ming.Single Frame Face Images Super-resolution Algorithm Based on WGAN[J].,2020,30(09):29-35.[doi:10. 3969 / j. issn. 1673-629X. 2020. 09. 006]
点击复制

基于 WGAN 单帧人脸图像超分辨率算法()
分享到：

《计算机技术与发展》[ISSN:1006-6977/CN:61-1281/TN]

卷:
30
期数:
2020年09期
页码:
29-35
栏目:
智能、算法、系统工程
出版日期:
2020-09-10

文章信息/Info

Title:
Single Frame Face Images Super-resolution Algorithm Based on WGAN
文章编号:
1673-629X(2020)09-0029-07
作者:
周传华1; 2; 吴幸运1; 李 鸣1
1. 安徽工业大学 管理科学与工程学院,安徽 马鞍山 243002; 2. 中国科学技术大学 计算机科学与技术学院,安徽 合肥 230026
Author(s):
ZHOU Chuan-hua1; 2; WU Xing-yun1; LI Ming1
1. School of Management Science & Engineering,Anhui University of Technology,Maanshan 243002,China; 2. School of Computer Science & Technology,University of Science & Technology of China,Hefei 230026,China
关键词:
生成对抗网络; Wasserstein 距离; 残差网络; 超分辨率重建; 深度学习
Keywords:
generative adversarial networks; Wasserstein distance; ResNet; super-resolution reconstruction; deep learning
分类号:
TP391. 41
DOI:
10. 3969 / j. issn. 1673-629X. 2020. 09. 006
摘要:
针对人脸超分辨率重建中引入的先验知识不够丰富的问题,提出了一种基于 Wasserstein 生成对抗网络(WGAN)的人脸超分辨率重建算法。 模型包含生成网络和判别网络,生成网络去除批量规范层并增加残差块数量加深网络深度,判别网络增加特征图通道数并引入了快捷连接优化网络,模型用 Wasserstein 距离代替 KL 散度作为网络的对抗损失,交替训练生成网络和判别网络,生成高分辨率的人脸图像。 实验结果表明,相比原始生成对抗网络超分辨率重建算法(SRGAN),所提算法在 MS-Celeb-1M 和 LFW 数据集中峰值信噪比(PSNR)和结构相似性(SSIM)分别提高了0.26dB、2%和 0.31dB、3% ,同时对比最近邻(NN)、双三次插值(Bic)、基于卷积神经网络超分辨率重建(SRCNN)、SRGAN,所提算法在 LFW、MS-Celeb-1M 数据集上均重建出视觉效果更好的人脸图像,证明了该算法的有效性,为人脸超分辨率重建提出了新的解决方案。
Abstract:
Aiming at lack of prior knowledge in the super-resolution reconstruction of face images,we propose a face super-resolution method based on Wasserstein generative adversarial network (WGAN). A generative net and adversarial net are designed in the model. For generative net, the batch-normalization layer of the standard residual block is removed and the number of residual block is increased to deepen network depth. Also,the number of feature maps is increased and skip connection is used in the discriminative to improve the network performance. The model uses Wasserstein distance instead of KL divergence as network adversarial loss. Alternate training generation network and discriminant network to generate high-quality face images. The experiment shows that compared with SRGAN, the improved algorithm has an increase of 0.26dB, 2% and 0.31 dB,3% in peak-signal-to-noise ratio (PSNR) and structural similarity (SSIM) in MS-Celeb-1M and LFW datasets. Compared with nearest neighbor (NN)、Bicubic、 SRCNN、 SRGAN, the improved algorithm can reconstruct the face images with better visual effect. It is proved that the proposed algorithm can effectively improve the resolution of face images and provides a new solution for face super-resolution reconstruction.

相似文献/References:

[1]康嘉钰,苏凡军.基于生成对抗网络的长短兴趣推荐模型[J].计算机技术与发展,2020,30(06):35.[doi:10. 3969 / j. issn. 1673-629X. 2020. 06. 007]
　KANG Jia-yu,SU Fan-jun.A Long-short-term Interests Recommendation Model Based on Generative Adversarial Networks[J].,2020,30(09):35.[doi:10. 3969 / j. issn. 1673-629X. 2020. 06. 007]
[2]蒋文杰,罗晓曙*,戴沁璇.一种改进的生成对抗网络的图像上色方法研究[J].计算机技术与发展,2020,30(07):56.[doi:10. 3969 / j. issn. 1673-629X. 2020. 07. 013]
　JIANG Wen-jie,LUO Xiao-shu*,DAI Qin-xuan.Research on an Improved Method of Generative Adversarial Networks Image Coloring[J].,2020,30(09):56.[doi:10. 3969 / j. issn. 1673-629X. 2020. 07. 013]
[3]庄兴旺,丁岳伟.多维度注意力和语义再生的文本生成图像模型[J].计算机技术与发展,2020,30(12):27.[doi:10. 3969 / j. issn. 1673-629X. 2020. 12. 005]
　ZHUANG Xing-wang,DING Yue-wei.Text-to-image Model by Multidimensional Attention and Semantic Regeneration[J].,2020,30(09):27.[doi:10. 3969 / j. issn. 1673-629X. 2020. 12. 005]
[4]尹玉婷,肖秦琨.基于深度卷积生成对抗网络的图像生成[J].计算机技术与发展,2021,31(04):86.[doi:10. 3969 / j. issn. 1673-629X. 2021. 04. 015]
　YIN Yu-ting,XIAO Qin-kun.Image Generation Based on Deep Convolution GenerativeAdversarial Networks[J].,2021,31(09):86.[doi:10. 3969 / j. issn. 1673-629X. 2021. 04. 015]
[5]程换新,张志浩,刘文翰,等.基于生成对抗网络的图像识别[J].计算机技术与发展,2021,31(06):175.[doi:10. 3969 / j. issn. 1673-629X. 2021. 06. 031]
　CHENG Huan-xin,ZHANG Zhi-hao,LIU Wen-han,et al.Image Recognition Based on Generative Adversarial Network[J].,2021,31(09):175.[doi:10. 3969 / j. issn. 1673-629X. 2021. 06. 031]
[6]苑金辉,乔 艳,费烨琳,等.基于深度迁移学习的心脏 MRI 图像左心室分割[J].计算机技术与发展,2021,31(06):35.[doi:10. 3969 / j. issn. 1673-629X. 2021. 06. 007]
　YUAN Jin-hui,QIAO Yan,FEI Ye-lin,et al.Left Ventricular Segmentation in Cardiac MRI Images Based onDeep Transfer Learning[J].,2021,31(09):35.[doi:10. 3969 / j. issn. 1673-629X. 2021. 06. 007]
[7]王田丰,胡谷雨,王 睿,等.基于 AAE 的网络性能异常发现[J].计算机技术与发展,2021,31(07):113.[doi:10. 3969 / j. issn. 1673-629X. 2021. 07. 019]
　WANG Tian-feng,HU Gu-yu,WANG Rui,et al.AAE-based Anomaly Detection for Network Performance[J].,2021,31(09):113.[doi:10. 3969 / j. issn. 1673-629X. 2021. 07. 019]
[8]徐志鹏,卢官明,罗燕晴.基于 CycleGAN 的人脸素描图像生成[J].计算机技术与发展,2021,31(08):63.[doi:10. 3969 / j. issn. 1673-629X. 2021. 08. 011]
　XU Zhi-peng,LU Guan-ming,LUO Yan-qing.Face Sketch Image Generation Based on CycleGAN[J].,2021,31(09):63.[doi:10. 3969 / j. issn. 1673-629X. 2021. 08. 011]
[9]王荣达,刘宁钟,李强懿,等.一种基于生成对抗网络的轻量级图像翻译模型[J].计算机技术与发展,2021,31(11):52.[doi:10. 3969 / j. issn. 1673-629X. 2021. 11. 009]
　WANG Rong-da,LIU Ning-zhong,LI Qiang-yi,et al.A Lightweight Image-to-image Translation Model Based on GAN[J].,2021,31(09):52.[doi:10. 3969 / j. issn. 1673-629X. 2021. 11. 009]
[10]植炜基,刘春雨,郑婉君,等.基于生成对抗网络的人脸表情识别技术综述[J].计算机技术与发展,2021,31(增刊):1.[doi:10. 3969 / j. issn. 1673-629X. 2021. S. 001]
　ZHI Wei-ji,LIU Chun-yu,ZHENG Wan-jun,et al.Survey of Facial Expression Recognition Technology Based onGenerative Adversarial Network[J].,2021,31(09):1.[doi:10. 3969 / j. issn. 1673-629X. 2021. S. 001]

更新日期/Last Update: 2020-09-10