Fragments Inpainting for Tomb Murals Using a Dual-Attention Mechanism GAN with Improved Generators

被引:4
|
作者
Wu, Meng [1 ,2 ]
Chang, Xiao [1 ]
Wang, Jia [3 ]
机构
[1] Xian Univ Architecture & Technol, Sch Informat & Control Engn, Xian 710055, Peoples R China
[2] Xian Univ Architecture & Technol, Inst Interdisciplinary & Innovate Res, Xian 710055, Peoples R China
[3] Shaanxi Hist Museum, Xian 710061, Peoples R China
来源
APPLIED SCIENCES-BASEL | 2023年 / 13卷 / 06期
基金
中国国家自然科学基金;
关键词
multiscale feature aggregation; double attention; GAN; image inpainting; tomb mural; RESTORATION; IMAGE; PAINTINGS; NETWORK;
D O I
10.3390/app13063972
中图分类号
O6 [化学];
学科分类号
0703 ;
摘要
As the only underground mural in the collection, the tomb murals are subject to damage due to temperature, humidity, and foundation settlement changes. Traditional mural inpainting takes a long time and requires experts to draw it manually. Therefore, the need for digital inpainting is increasing to save time and costs. Due to the scarcity of samples and the variety of damage, the image features are scattered and partially sparse, and the colors are less vivid than in other images. Traditional deep learning inpainting causes information loss and generates irrational structures. The generative adversarial network is, recently, a more effective method. Therefore, this paper presents an inpainting model based on dual-attention multiscale feature aggregation and an improved generator. Firstly, an improved residual prior and attention mechanism is added to the generator module to preserve the image structure. Secondly, the model combines spatial and channel attention with multiscale feature aggregation to change the mapping network structure and improve the inpainting accuracy. Finally, the segmental loss function and its training method are improved.The experimental results show that the results of using signal-to-noise ratio (PSNR), structural similarity (SSIM), and mean square error (MSE) on epitaxial mask, crack mask, random small mask, and random large mask are better than other methods. It demonstrates the performance of this paper in inpainting different diseases of murals. It can be used as a reference for experts in manual inpainting, saving the cost and time of manual inpainting.
引用
收藏
页数:19
相关论文
共 50 条
  • [31] Social Network Rumor Detection Method Combining Dual-Attention Mechanism With Graph Convolutional Network
    Liu, Xiaoyang
    Zhao, Zhengyang
    Zhang, Yihao
    Liu, Chao
    Yang, Fan
    IEEE TRANSACTIONS ON COMPUTATIONAL SOCIAL SYSTEMS, 2023, 10 (05) : 2350 - 2361
  • [32] A Dual-Attention Mechanism Deep Learning Network for Mesoscale Eddy Detection by Mining Spatiotemporal Characteristics
    Li, Baixin
    Tang, Huan
    Ma, Dongfang
    Lin, Jianmin
    Journal of Atmospheric and Oceanic Technology, 2022, 39 (08): : 1115 - 1128
  • [33] DA-GAN: Dual-attention generative adversarial networks for real-world exquisite makeup transfer
    Jiao, Qianfen
    Xu, Zhen
    Wu, Si
    Wong, Hau-San
    PATTERN RECOGNITION, 2025, 158
  • [34] Automatic Prediction of Multiple Associated Diseases Using a Dual-Attention Neural Network Model
    Ren, Yafeng
    Wang, Zilin
    Tong, Andwei
    HEALTH INFORMATION PROCESSING, CHIP 2023, 2023, 1993 : 369 - 391
  • [35] A pyramid auxiliary supervised U-Net model for road crack detection with dual-attention mechanism
    Lu, Yingxiang
    Zhang, Guangyuan
    Duan, Shukai
    Chen, Feng
    DISPLAYS, 2024, 84
  • [36] Hybrid modeling for vehicle lateral dynamics via AGRU with a dual-attention mechanism under limited data
    Chen, Jianwei
    Yu, Chuanqiang
    Wang, Yafei
    Zhou, Zhisong
    Liu, Zhihao
    CONTROL ENGINEERING PRACTICE, 2024, 151
  • [37] A Generative Adversarial Network Fused with Dual-Attention Mechanism and Its Application in Multitarget Image Fine Segmentation
    Yin, Jian
    Zhou, Zhibo
    Xu, Shaohua
    Yang, Ruiping
    Liu, Kun
    COMPUTATIONAL INTELLIGENCE AND NEUROSCIENCE, 2021, 2021
  • [38] End-to-end model-based trajectory prediction for ro-ro ship route using dual-attention mechanism
    Zhao, Licheng
    Zuo, Yi
    Zhang, Wenjun
    Li, Tieshan
    Chen, C. L. Philip
    FRONTIERS IN COMPUTATIONAL NEUROSCIENCE, 2024, 18
  • [39] Prediction of the remaining useful life of rolling bearings by LSTM based on multidomain characteristics and a dual-attention mechanism
    Bao, Huaiqian
    Song, Lijin
    Zhang, Zongzhen
    Han, Baokun
    Wang, Jinrui
    Ma, Junqing
    Jiang, Xingwang
    JOURNAL OF MECHANICAL SCIENCE AND TECHNOLOGY, 2023, 37 (09) : 4583 - 4596
  • [40] Prediction of the remaining useful life of rolling bearings by LSTM based on multidomain characteristics and a dual-attention mechanism
    Huaiqian Bao
    Lijin Song
    Zongzhen Zhang
    Baokun Han
    Jinrui Wang
    Junqing Ma
    Xingwang Jiang
    Journal of Mechanical Science and Technology, 2023, 37 : 4583 - 4596