TY - JOUR
T1 - GCN-Denoiser
T2 - Mesh Denoising with Graph Convolutional Networks
AU - Shen, Yuefan
AU - Fu, Hongbo
AU - Du, Zhongshuo
AU - Chen, Xiang
AU - Burnaev, Evgeny
AU - Zorin, Denis
AU - Zhou, Kun
AU - Zheng, Youyi
N1 - Funding Information:
This work was supported in part by the National Key Research & Development Program of China (2018YFE0100900) and the NSF China (Grants No. 61890954, No. 61772024, and No. 61732016). Denis Zorin was supported by the Ministry of Education and Science of Russian Federation, grant No. 14.615.21.0004, grant code: RFMEFI61518X0004. Evgeny Burnaev was supported by the Ministry of Science and Higher Education grant No. 075-10-2021-068. Authors’ addresses: Y. Shen, The State Key Lab of CAD&CG, Zhejiang University, 866 Yuhangtang Rd, Hangzhou, China; email: [email protected]; H. Fu, The School of Creative Media, City University of Hong Kong; email: [email protected]; Z. Du and X. Chen, The State Key Lab of CAD&CG, Zhejiang University; emails: [email protected], [email protected]; E. Burnaev, Skolkovo Institute of Science and Technology, Moscow, Russia; email: [email protected]; D. Zorin, New York University, New York, The United States; email: [email protected]; K. Zhou, The State Key Lab of CAD&CG, Zhejiang University; email: [email protected]; Y. Zheng, The State Key Lab of CAD&CG, Zhejiang University; email: [email protected].
Publisher Copyright:
© 2022 Copyright held by the owner/author(s). Publication rights licensed to ACM.
PY - 2022/2
Y1 - 2022/2
N2 - In this article, we present GCN-Denoiser, a novel feature-preserving mesh denoising method based on graph convolutional networks (GCNs). Unlike previous learning-based mesh denoising methods that exploit handcrafted or voxel-based representations for feature learning, our method explores the structure of a triangular mesh itself and introduces a graph representation followed by graph convolution operations in the dual space of triangles. We show such a graph representation naturally captures the geometry features while being lightweight for both training and inference. To facilitate effective feature learning, our network exploits both static and dynamic edge convolutions, which allow us to learn information from both the explicit mesh structure and potential implicit relations among unconnected neighbors. To better approximate an unknown noise function, we introduce a cascaded optimization paradigm to progressively regress the noise-free facet normals with multiple GCNs. GCN-Denoiser achieves the new state-of-the-art results in multiple noise datasets, including CAD models often containing sharp features and raw scan models with real noise captured from different devices. We also create a new dataset called PrintData containing 20 real scans with their corresponding ground-truth meshes for the research community. Our code and data are available at https://github.com/Jhonve/GCN-Denoiser.
AB - In this article, we present GCN-Denoiser, a novel feature-preserving mesh denoising method based on graph convolutional networks (GCNs). Unlike previous learning-based mesh denoising methods that exploit handcrafted or voxel-based representations for feature learning, our method explores the structure of a triangular mesh itself and introduces a graph representation followed by graph convolution operations in the dual space of triangles. We show such a graph representation naturally captures the geometry features while being lightweight for both training and inference. To facilitate effective feature learning, our network exploits both static and dynamic edge convolutions, which allow us to learn information from both the explicit mesh structure and potential implicit relations among unconnected neighbors. To better approximate an unknown noise function, we introduce a cascaded optimization paradigm to progressively regress the noise-free facet normals with multiple GCNs. GCN-Denoiser achieves the new state-of-the-art results in multiple noise datasets, including CAD models often containing sharp features and raw scan models with real noise captured from different devices. We also create a new dataset called PrintData containing 20 real scans with their corresponding ground-truth meshes for the research community. Our code and data are available at https://github.com/Jhonve/GCN-Denoiser.
KW - Mesh denoising
KW - cascaded optimization
KW - graph convolutional networks
UR - http://www.scopus.com/inward/record.url?scp=85124798401&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85124798401&partnerID=8YFLogxK
U2 - 10.1145/3480168
DO - 10.1145/3480168
M3 - Article
AN - SCOPUS:85124798401
SN - 0730-0301
VL - 41
JO - ACM Transactions on Graphics
JF - ACM Transactions on Graphics
IS - 1
M1 - 8
ER -