CageNet: A Meta-Framework for Learning on Wild Meshes
Not on the map. Why?
Abstract (arXiv)
Learning on triangle meshes has recently proven to be instrumental to a myriad of tasks, from shape classification, to segmentation, to deformation and animation, to mention just a few. While some of these applications are tackled through neural network architectures which are tailored to the application at hand, many others use generic frameworks for triangle meshes where the only customization required is the modification of the input features and the loss function. Our goal in this paper is to broaden the applicability of these generic frameworks to "wild", i.e. meshes in-the-wild which often have multiple components, non-manifold elements, disrupted connectivity, or a combination of these. We propose a configurable meta-framework based on the concept of caged geometry: Given a mesh, a cage is a single component manifold triangle mesh that envelopes it closely. Generalized barycentric coordinates map between functions on the cage, and functions on the mesh, allowing us to learn and test on a variety of data, in different applications. We demonstrate this concept by learning segmentation and skinning weights on difficult data, achieving better performance to state of the art techniques on wild meshes.
Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Builds on 30 of its 76 references that are in the map. First posted on arXiv 24 May 2025.
Builds on
30 of 76 references.
References · 30
- 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,305
- 2016PointNet: Deep Learning on Point Sets for 3D Classification and Segmentationcited by 5,248
- 2015ShapeNet: An Information-Rich 3D Model Repositorycited by 2,897
- 2018Dynamic Graph CNN for Learning on Point Cloudscited by 2,223
- 2016Geometric deep learning: going beyond Euclidean datacited by 1,048
- 2024Structured 3D Latents for Scalable and Versatile 3D Generationcited by 770
- 2020DiffusionNet: Discretization Agnostic Learning on Surfacescited by 164
- 2016Learning shape correspondence with anisotropic convolutional neural networkscited by 95
- 2019SpiralNet++: A Fast and Highly Efficient Mesh Convolution Operatorcited by 89
- 2022Deforming Radiance Fields with Cagescited by 78
- 2022Dual Octree Graph Networks for Learning Adaptive Volumetric Shape Representationscited by 71
- 2018MeshCNN: A Network with an Edgecited by 69
18 more
- 2021Subdivision-Based Mesh Convolution Networkscited by 60
- 2020Deep Geometric Texture Synthesiscited by 51
- 2021KeypointDeformer: Unsupervised 3D Keypoint Discovery for Shape Controlcited by 51
- 2020Neural Subdivisioncited by 49
- 2020Primal-Dual Mesh Convolutional Neural Networkscited by 44
- 2020CNNs on Surfaces using Rotation-Equivariant Featurescited by 43
- 2018A Simple Approach to Intrinsic Correspondence Learning on Unstructured 3D Meshescited by 39
- 2020ManifoldPlus: A Robust and Scalable Watertight Manifold Surface Generation Method for Triangle Soupscited by 38
- 2019Fast Tetrahedral Meshing in the Wildcited by 35
- 2018Multi-directional Geodesic Neural Networks via Equivariant Convolutioncited by 32
- 2020RigNet: Neural Rigging for Articulated Characterscited by 32
- 2020MeshWalker: Deep Mesh Understanding by Random Walkscited by 32
- 2018Surface Networks via General Coverscited by 22
- 2022SkinningNet: Two-Stream Graph Convolutional Neural Network for Skinning Prediction of Synthetic Characterscited by 16
- 2020Continuous Geodesic Convolutions for Learning on 3D Shapescited by 14
- 2021HodgeNet: Learning Spectral Geometry on Triangle Meshescited by 13
- 2021HeterSkinNet: A Heterogeneous Network for Skin Weights Predictioncited by 5
- 2023SieveNet: Selecting Point-Based Features for Mesh Networkscited by 1
Built on by
Citing papers per year: 2026: 1.
1 paper.