Hierarchical CADNet: Learning from B-Reps for Machining Feature Recognition

Andrew Colligan; Trevor T Robinson; Declan C. Nolan; Yang Hua; Weijuan Cao

doi:10.1016/j.cad.2022.103226

Hierarchical CADNet: Learning from B-Reps for Machining Feature Recognition

Andrew Colligan, Trevor T Robinson, Declan C. Nolan, Yang Hua, Weijuan Cao

Research output: Contribution to journal › Article › peer-review

235 Downloads (Pure)

Abstract

Deep learning approaches have been shown to be capable of recognizing shape features (e.g. machining features) in Computer-Aided Design (CAD) models in certain circumstances, yet still have issues when the features intersect, and in exploiting the geometric and topological information which comprises the boundary representation (B-Rep) of the typical CAD model. This paper presents a novel hierarchical B-Rep graph shape representation which encodes information about the surface geometry and face topology of the B-Rep. To learn from this new shape representation, a novel hierarchical graph convolutional network called Hierarchical CADNet has been created, which has been shown to outperform other state-of-the-art neural architectures on feature identification, including machining features that intersect, with improvements in accuracy for some more complex CAD models.

Original language	English
Article number	103226
Number of pages	16
Journal	CAD Computer Aided Design
Volume	147
Issue number	103226
Early online date	19 Feb 2022
DOIs	https://doi.org/10.1016/j.cad.2022.103226
Publication status	Published - 01 Jun 2022

Keywords

Machining feature recognition
3D deep learning
Hierarchical graph convolution network
Computer-aided process planning (CAPP)
B-Rep
CAD

Access to Document

10.1016/j.cad.2022.103226Licence: CC BY

Hierarchical CADNet: Learning from B-Reps for Machining Feature Recognition
Copyright 2022 the authors. This is an open access article published under a Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution and reproduction in any medium, provided the author and source are cited.
Final published version, 4.48 MBLicence: CC BY

MFCAD++ Dataset. Dataset for paper: "Hierarchical CADNet: Learning from B-Reps for Machining Feature Recognition, Computer-Aided Design"
Colligan, A. (Creator), Robinson, T. T. (Creator), Nolan, D. (Creator) & Hua, Y. (Creator), Queen's University Belfast, 12 Feb 2022
DOI: 10.17034/d1fec5a0-8c10-4630-b02e-b92dc81df823
Dataset

File
Hierarchical-CADNet neural network code
Colligan, A. (Creator), Queen's University Belfast, May 2022
https://gitlab.com/qub_femg/machine-learning/hierarchical-cadnet
Dataset
Hierarchical B-Rep graph code
Colligan, A. (Creator), Queen's University Belfast, May 2022
https://gitlab.com/qub_femg/machine-learning/hierarchical-brep-graphs
Dataset

Deep learning for boundary representation CAD models
Author: Colligan, A., Jul 2022
Supervisor: Nolan, D. (Supervisor), Hua, Y. (Supervisor) & Robinson, T. (Supervisor)
Student thesis: Doctoral Thesis › Doctor of Philosophy

File

Cite this

@article{0aeb13bffe2b46a0907fd5ba499ab305,

title = "Hierarchical CADNet: Learning from B-Reps for Machining Feature Recognition",

abstract = "Deep learning approaches have been shown to be capable of recognizing shape features (e.g. machining features) in Computer-Aided Design (CAD) models in certain circumstances, yet still have issues when the features intersect, and in exploiting the geometric and topological information which comprises the boundary representation (B-Rep) of the typical CAD model. This paper presents a novel hierarchical B-Rep graph shape representation which encodes information about the surface geometry and face topology of the B-Rep. To learn from this new shape representation, a novel hierarchical graph convolutional network called Hierarchical CADNet has been created, which has been shown to outperform other state-of-the-art neural architectures on feature identification, including machining features that intersect, with improvements in accuracy for some more complex CAD models.",

keywords = "Machining feature recognition, 3D deep learning, Hierarchical graph convolution network, Computer-aided process planning (CAPP), B-Rep, CAD",

author = "Andrew Colligan and Robinson, {Trevor T} and Nolan, {Declan C.} and Yang Hua and Weijuan Cao",

year = "2022",

month = jun,

day = "1",

doi = "10.1016/j.cad.2022.103226",

language = "English",

volume = "147",

journal = "Computer-Aided Design",

issn = "0010-4485",

publisher = "Elsevier Limited",

number = "103226",

}

TY - JOUR

T1 - Hierarchical CADNet: Learning from B-Reps for Machining Feature Recognition

AU - Colligan, Andrew

AU - Robinson, Trevor T

AU - Nolan, Declan C.

AU - Hua, Yang

AU - Cao, Weijuan

PY - 2022/6/1

Y1 - 2022/6/1

N2 - Deep learning approaches have been shown to be capable of recognizing shape features (e.g. machining features) in Computer-Aided Design (CAD) models in certain circumstances, yet still have issues when the features intersect, and in exploiting the geometric and topological information which comprises the boundary representation (B-Rep) of the typical CAD model. This paper presents a novel hierarchical B-Rep graph shape representation which encodes information about the surface geometry and face topology of the B-Rep. To learn from this new shape representation, a novel hierarchical graph convolutional network called Hierarchical CADNet has been created, which has been shown to outperform other state-of-the-art neural architectures on feature identification, including machining features that intersect, with improvements in accuracy for some more complex CAD models.

AB - Deep learning approaches have been shown to be capable of recognizing shape features (e.g. machining features) in Computer-Aided Design (CAD) models in certain circumstances, yet still have issues when the features intersect, and in exploiting the geometric and topological information which comprises the boundary representation (B-Rep) of the typical CAD model. This paper presents a novel hierarchical B-Rep graph shape representation which encodes information about the surface geometry and face topology of the B-Rep. To learn from this new shape representation, a novel hierarchical graph convolutional network called Hierarchical CADNet has been created, which has been shown to outperform other state-of-the-art neural architectures on feature identification, including machining features that intersect, with improvements in accuracy for some more complex CAD models.

KW - Machining feature recognition

KW - 3D deep learning

KW - Hierarchical graph convolution network

KW - Computer-aided process planning (CAPP)

KW - B-Rep

KW - CAD

U2 - 10.1016/j.cad.2022.103226

DO - 10.1016/j.cad.2022.103226

M3 - Article

VL - 147

JO - Computer-Aided Design

JF - Computer-Aided Design

SN - 0010-4485

IS - 103226

M1 - 103226

ER -

Hierarchical CADNet: Learning from B-Reps for Machining Feature Recognition

Abstract

Keywords

Access to Document

Fingerprint

Datasets

MFCAD++ Dataset. Dataset for paper: "Hierarchical CADNet: Learning from B-Reps for Machining Feature Recognition, Computer-Aided Design"

Hierarchical-CADNet neural network code

Hierarchical B-Rep graph code

Student theses

Deep learning for boundary representation CAD models

Cite this