BRepNet: A topological message passing system for solid models

@article{Lambourne2021BRepNetAT,
  title={BRepNet: A topological message passing system for solid models},
  author={J. Lambourne and Karl D. D. Willis and Pradeep Kumar Jayaraman and Aditya Sanghi and Peter Meltzer and Hooman Shayani},
  journal={2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
  year={2021},
  pages={12768-12777}
}
Boundary representation (B-rep) models are the standard way 3D shapes are described in Computer-Aided Design (CAD) applications. They combine lightweight parametric curves and surfaces with topological information which connects the geometric entities to describe manifolds. In this paper we introduce BRepNet, a neural network architecture designed to operate directly on B-rep data structures, avoiding the need to approximate the model as meshes or point clouds. BRepNet defines convolutional… 

Figures and Tables from this paper

3D Adversarial Attacks Beyond Point Cloud
TLDR
This paper proposes a strong 3D adversarial attack named Mesh Attack to address this problem by directly performing perturbation on mesh of a 3D object and demonstrates that the proposed scheme outperforms SOTA 3D attacks by a significant margin.
SB-GCN: Structured BREP Graph Convolutional Network for Automatic Mating of CAD Assemblies
TLDR
The authors' tool assists a designer in assembling a collection of CAD parts into a functional assembly by creating embeddings for each topological entity of the selected parts’ BREPs, which are used as input to a classifier that scores potential mates defined in reference to these topological entities.
ComplexGen: CAD Reconstruction by B-Rep Chain Complex Generation
TLDR
A novel neural framework is proposed that consists of a sparse CNN encoder for input point cloud processing and a tri-path transformer decoder for generating geometric primitives and their mutual relationships with estimated probabilities, leading to structurally more faithful and complete CAD B-Rep models than previous results.
Mates2Motion: Learning How Mechanical CAD Assemblies Work
TLDR
This work describes the work on inferring the degrees of freedom between mated parts in mechanical assemblies using deep learning on CAD representations, and presents methods for re-defining these mates to make them better re-deen the motion of the assembly, as well as narrowing down the possible axes of motion.
RendNet: Unified 2D/3D Recognizer With Latent Space Rendering
TLDR
RendNet is proposed, a unified architecture for recognition on both 2D and 3D scenarios, which considers both VG/RG representations and exploits their interaction by incorporating the VG-to-RG rasterization process.
SolidGen: An Autoregressive Model for Direct B-rep Synthesis
TLDR
This paper presents a new approach that enables learning from and synthesizing B-reps without the need for supervision through CAD modeling sequence data, and demonstrates qualitatively, quantitatively and through perceptual evaluation by human subjects that SolidGen can produce high quality, realistic looking CAD models.
Neural Face Identification in a 2D Wireframe Projection of a Manifold Object
TLDR
This paper approaches the classical problem of face identification from a novel data-driven point of view, and adopts a variant of the popular Transformer model to predict the edges associated with the same face in a natural order.
Point2Cyl: Reverse Engineering 3D Objects from Point Clouds to Extrusion Cylinders
TLDR
A neural network is introduced that solves the extrusion cylinder decomposition problem in a geometry-grounded way by first learning underlying geometric proxies, and predicts per-point segmentation, base/barrel labels and normals, then estimates for the underlying extrusion parameters in differentiable and closed-form formulations.
JoinABLe: Learning Bottom-up Assembly of Parametric CAD Joints
TLDR
This paper introduces JoinABLe, a learning-based method that assembles parts together to form joints without the help of object class labels or human guidance, and releases the Fusion 360 Gallery assembly dataset, containing assemblies with rich information on joints, contact surfaces, holes, and the underlying assembly graph structure.
Constraint Synthesis for Parametric CAD
TLDR
This work synthesizes constraints on parameters of parametric designs statically, by exploiting implicit assumptions of the design process, with a very small runtime cost, in the order of seconds.
...
...

References

SHOWING 1-10 OF 54 REFERENCES
Neural 3D Morphable Models: Spiral Convolutional Networks for 3D Shape Representation Learning and Generation
TLDR
This work introduces a novel graph convolutional operator, acting directly on the 3D mesh, that explicitly models the inductive bias of the fixed underlying graph, by enforcing consistent local orderings of the vertices of the graph, through the spiral operator.
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation
TLDR
This paper designs a novel type of neural network that directly consumes point clouds, which well respects the permutation invariance of points in the input and provides a unified architecture for applications ranging from object classification, part segmentation, to scene semantic parsing.
ABC: A Big CAD Model Dataset for Geometric Deep Learning
TLDR
This work performs a large-scale benchmark for estimation of surface normals, comparing existing data driven methods and evaluating their performance against both the ground truth and traditional normal estimation methods.
Neural subdivision
This paper introduces Neural Subdivision, a novel framework for data-driven coarse-to-fine geometry modeling. During inference, our method takes a coarse triangle mesh as input and recursively
Dynamic Graph CNN for Learning on Point Clouds
TLDR
This work proposes a new neural network module suitable for CNN-based high-level tasks on point clouds, including classification and segmentation called EdgeConv, which acts on graphs dynamically computed in each layer of the network.
PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Space
TLDR
A hierarchical neural network that applies PointNet recursively on a nested partitioning of the input point set and proposes novel set learning layers to adaptively combine features from multiple scales to learn deep point set features efficiently and robustly.
Topological Structures for Geometric Modeling
TLDR
A non-two-manifold boundary geometric modeling topology representation is developed which allows the unified and simultaneous representation of wireframe, surface, and solid modeling forms, while featuring a representable range beyond what is achievable in any of the previous modeling forms.
Geometric Modeling Using the Euler Operators
TLDR
This paper offers a description of the Euler operators, in a form expected to be useful for prospective implementers and others wishing to better understand their function and behavior.
A Deep 3D Convolutional Neural Network Based Design for Manufacturability Framework
TLDR
A framework using Deep 3D-CNNs to learn salient features from a CAD model of a mechanical part and determine if the part can be manufactured or not and is able to learn the DFM rules for non-manufacturable features without a human prior is developed.
...
...