Deep 3D mesh watermarking with self-adaptive robustness

Cybersecurity

Research

Han Fang¹  Hang Zhou²  Feng Wang³  Nenghai Yu³  Weiming Zhang³ 

[1] National University of Singapore, Kent Ridge, Singapore;Simon Fraser University, Vancouver, Canada;University of Science and Technology of China, Hefei, China;

关键词: 3D mesh watermarking;    Graph convolution network;    Attack layer;   

DOI  :  10.1186/s42400-022-00125-w

 received in 2022-01-31, accepted in 2022-04-17,  发布年份 2022

来源: Springer

【 摘 要 】

Robust 3D mesh watermarking is a traditional research topic in computer graphics, which provides an efficient solution to the copyright protection for 3D meshes. Traditionally, researchers need manually design watermarking algorithms to achieve sufficient robustness for the actual application scenarios. In this paper, we propose the first deep learning-based 3D mesh watermarking network, which can provide a more general framework for this problem. In detail, we propose an end-to-end network, consisting of a watermark embedding sub-network, a watermark extracting sub-network and attack layers. We employ the topology-agnostic graph convolutional network (GCN) as the basic convolution operation, therefore our network is not limited by registered meshes (which share a fixed topology). For the specific application scenario, we can integrate the corresponding attack layers to guarantee adaptive robustness against possible attacks. To ensure the visual quality of watermarked 3D meshes, we design the curvature consistency loss function to constrain the local geometry smoothness of watermarked meshes. Experimental results show that the proposed method can achieve more universal robustness while guaranteeing comparable visual quality.

【 授权许可】

CC BY   
© The Author(s) 2022

【 图 表 】

Fig. 2

Fig. 8

Fig. 1

Fig. 2

Fig. 5

Fig. 2

Fig. 1

Fig. 5

Fig. 3

Fig. 2

12864_2022_9026_Article_IEq102.gif

Fig. 2

Fig. 2

Fig. 8

Fig. 2

Fig. 1

Fig. 1

Fig. 2

Fig. 4

Fig. 3

Fig. 3

Fig. 2

Fig. 1

12864_2022_9026_Article_IEq78.gif

Fig. 1

Fig. 2

Fig. 2

Fig. 1

Fig. 1

Fig. 2

Fig. 3

Fig. 1

Fig. 4

Fig. 1

Scheme 1

Fig. 2

Fig. 1

Fig. 1

Fig. 4

Fig. 1

Fig. 2

Fig. 1

Fig. 1

Fig. 1

Fig.1

Fig. 1

Fig. 6

12864_2022_9026_Article_IEq59.gif

Fig. 1

12864_2022_9026_Article_IEq57.gif

Fig. 1

12864_2022_9026_Article_IEq55.gif

Fig. 3

Fig. 1

Fig. 4

Fig. 3

Fig. 6

41408_2022_764_Article_IEq31.gif

41408_2022_764_Article_IEq30.gif

41408_2022_764_Article_IEq28.gif

Fig. 1

41408_2022_764_Article_IEq26.gif

41408_2022_764_Article_IEq25.gif

41408_2022_764_Article_IEq24.gif

41408_2022_764_Article_IEq23.gif

41408_2022_764_Article_IEq22.gif

41408_2022_764_Article_IEq21.gif

41408_2022_764_Article_IEq20.gif

41408_2022_764_Article_IEq19.gif

Fig. 1

41408_2022_764_Article_IEq17.gif

41408_2022_764_Article_IEq16.gif

41408_2022_764_Article_IEq15.gif

41408_2022_764_Article_IEq14.gif

41408_2022_764_Article_IEq13.gif

41408_2022_764_Article_IEq12.gif

41408_2022_764_Article_IEq11.gif

41408_2022_764_Article_IEq10.gif

41408_2022_764_Article_IEq9.gif

41408_2022_764_Article_IEq8.gif

41408_2022_764_Article_IEq7.gif

41408_2022_764_Article_IEq6.gif

Fig. 2

41408_2022_764_Article_IEq3.gif

Fig. 1

41408_2022_764_Article_IEq1.gif

Fig. 2

Fig. 5

Fig. 1

40249_2022_1045_Article_IEq21.gif

Fig. 2

Fig. 2

Fig. 1

Fig. 1

Fig. 1

Fig. 2

Fig. 1

Fig. 1

Fig. 1