Journal
ACM TRANSACTIONS ON GRAPHICS
Volume 41, Issue 3, Pages -Publisher
ASSOC COMPUTING MACHINERY
DOI: 10.1145/3506694
Keywords
Geometric deep learning; convolutional neural network; subdivision surfaces; mesh processing
Categories
Funding
- Natural Science Foundation of China [61521002]
- Research Grant of Beijing Higher Institution Engineering Research Center
- Tsinghua-Tencent Joint Laboratory for Internet Innovation Technology
Ask authors/readers for more resources
This article presents SubdivNet, an innovative and versatile CNN framework for processing three-dimensional triangle meshes. By utilizing the subdivision structure of meshes and making an analogy to 2D images, SubdivNet enables convolution and pooling operations on meshes, making it possible to adapt popular 2D CNN architectures for 3D mesh processing. The approach allows remeshing of meshes with arbitrary connectivity to have Loop subdivision sequence connectivity via self-parameterization.
Convolutionalneural networks (CNNs) have made great breakthroughs in two-dimensional (2D) computer vision. However, their irregular structure makes it hard to harness the potential of CNNs directly on meshes. A subdivision surface provides a hierarchical multi-resolution structure in which each face in a closed 2-manifold triangle mesh is exactly adjacent to three faces. Motivated by these two observations, this article presents SubdivNet, an innovative and versatile CNN framework for three-dimensional (3D) triangle meshes with Loop subdivision sequence connectivity. Making an analogy between mesh faces and pixels in a 2D image allows us to present a mesh convolution operator to aggregate local features from nearby faces. By exploiting face neighborhoods, this convolution can support standard 2D convolutional network concepts, e.g., variable kernel size, stride, and dilation. Based on the multi-resolution hierarchy, we make use of pooling layers that uniformly merge four faces into one and an upsampling method that splits one face into four. Thereby, many popular 2D CNN architectures can be easily adapted to process 3D meshes. Meshes with arbitrary connectivity can be remeshed to have Loop subdivision sequence connectivity via self-parameterization, making SubdivNet a general approach. Extensive evaluation and various applications demonstrate SubdivNet's effectiveness and efficiency.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available