An image watermark removal method for secure internet of things applications based on federated learning

被引:3
作者
Li, Hongan [1 ]
Wang, Guanyi [1 ]
Hua, Qiaozhi [2 ]
Wen, Zheng [3 ]
Li, Zhanli [1 ]
Lei, Ting [1 ]
机构
[1] Xian Univ Sci & Technol, Coll Comp Sci & Technol, Xian, Peoples R China
[2] Hubei Univ Arts & Sci, Comp Sch, Xiangyang, Hubei, Peoples R China
[3] Waseda Univ, Sch Fundamental Sci & Engn, Waseda, Japan
关键词
deep image prior; federated learning; image security and privacy protection; image watermark removal; internet of things; SCHEME;
D O I
10.1111/exsy.13036
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Watermark adding is one of the important means for image security and privacy protection in Internet of things (IOT) applications based on federated learning. It is often inseparable from adversarial training with watermark removal algorithms. The effect of watermark removal algorithms will directly affect the final result of watermark addition. However, the existing watermark removal algorithms have drawbacks such as incomplete image watermark removal, poor image quality after watermark removal, large demand for training data, and incorrect filling, which seriously affects the development of image information security and privacy protection in IOT applications based on federated learning. To solve the above problems, this paper proposes an improved image watermark removal convolutional network model based on deep image prior. First, we improve the U-Net network model, using six downsamping layers and six deconvolution layers combined with deep image prior method to reduce the loss of details and perceive high-level features, thereby improving the ability of the network to extract high-level features of the image. In addition, we design a new type of loss function which is called stair loss, and add L1 loss and perception loss to establish new constraints. In order to verify the effectiveness of our method, a comprehensive experimental comparison was conducted on the public dataset PASCAL VOC 2012 in the same experimental environment with CGAN and the deep prior method. The experimental results show that the improved model combined with the deep image prior method can extract the high-level feature information and can directly remove the watermark from the picture without pretraining the network, the L1 loss and perceptual loss can better retain the image structure information and speed up the watermark removal of the model, the stair loss corrects the final output more accurately by correcting the output of each layer; our method improves the learning ability of the model, and under the condition of the same training time, the image quality after watermark removal is higher, and the final watermark removal result is better, which is more suitable for distributed structure of IoT application based on federated learning.
引用
收藏
页数:15
相关论文
共 51 条
[1]   Federated Learning for Cybersecurity: Concepts, Challenges, and Future Directions [J].
Alazab, Mamoun ;
Priya, Swarna R. M. ;
Parimala, M. ;
Maddikunta, Praveen Kumar Reddy ;
Gadekallu, Thippa Reddy ;
Quoc-Viet Pham .
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS, 2022, 18 (05) :3501-3509
[2]  
Bo L., 2019, THESIS HARBIN I TECH
[3]   Generative adversarial networks model for visible watermark removal [J].
Cao, Zhiyi ;
Niu, Shaozhang ;
Zhang, Jiwei ;
Wang, Xinyi .
IET IMAGE PROCESSING, 2019, 13 (10) :1783-1789
[4]   Large-Scale Visible Watermark Detection and Removal with Deep Convolutional Networks [J].
Cheng, Danni ;
Li, Xiang ;
Li, Wei-Hong ;
Lu, Chan ;
Li, Fake ;
Zhao, Hua ;
Zheng, Wei-Shi .
PATTERN RECOGNITION AND COMPUTER VISION, PT III, 2018, 11258 :27-40
[5]   On the Effectiveness of Visible Watermarks [J].
Dekel, Tali ;
Rubinstein, Michael ;
Liu, Ce ;
Freeman, William T. .
30TH IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR 2017), 2017, :6864-6872
[6]   Perceptual Enhancement for Autonomous Vehicles: Restoring Visually Degraded Images for Context Prediction via Adversarial Training [J].
Ding, Feng ;
Yu, Keping ;
Gu, Zonghua ;
Li, Xiangjun ;
Shi, Yunqing .
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS, 2022, 23 (07) :9430-9441
[7]   Deep-Learning-Empowered Digital Forensics for Edge Consumer Electronics in 5G HetNets [J].
Ding, Feng ;
Zhu, Guopu ;
Alazab, Mamoun ;
Li, Xiangjun ;
Yu, Keping .
IEEE CONSUMER ELECTRONICS MAGAZINE, 2022, 11 (02) :42-50
[8]   Generating high quality crowd density map based on perceptual loss [J].
Fan, Zheyi ;
Zhu, Yixuan ;
Song, Yu ;
Liu, Zhiwen .
APPLIED INTELLIGENCE, 2020, 50 (04) :1073-1085
[9]   Blockchain-Based Cross-Domain Authentication for Intelligent 5G-Enabled Internet of Drones [J].
Feng, Chaosheng ;
Liu, Bin ;
Guo, Zhen ;
Yu, Keping ;
Qin, Zhiguang ;
Choo, Kim-Kwang Raymond .
IEEE INTERNET OF THINGS JOURNAL, 2022, 9 (08) :6224-6238
[10]   Blockchain-Empowered Decentralized Horizontal Federated Learning for 5G-Enabled UAVs [J].
Feng, Chaosheng ;
Liu, Bin ;
Yu, Keping ;
Goudos, Sotirios K. ;
Wan, Shaohua .
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS, 2022, 18 (05) :3582-3592