Show simple item record

dc.contributor.authorPara, Wamiq Reyaz
dc.contributor.authorBhat, Shariq Farooq
dc.contributor.authorGuerrero, Paul
dc.contributor.authorKelly, Tom
dc.contributor.authorMitra, Niloy
dc.contributor.authorGuibas, Leonidas
dc.contributor.authorWonka, Peter
dc.description.abstractComputer-aided design (CAD) is the most widely used modeling approach for technical design. The typical starting point in these designs is 2D sketches which can later be extruded and combined to obtain complex three-dimensional assemblies. Such sketches are typically composed of parametric primitives, such as points, lines, and circular arcs, augmented with geometric constraints linking the primitives, such as coincidence, parallelism, or orthogonality. Sketches can be represented as graphs, with the primitives as nodes and the constraints as edges. Training a model to automatically generate CAD sketches can enable several novel workflows, but is challenging due to the complexity of the graphs and the heterogeneity of the primitives and constraints. In particular, each type of primitive and constraint may require a record of different size and parameter types. We propose SketchGen as a generative model based on a transformer architecture to address the heterogeneity problem by carefully designing a sequential language for the primitives and constraints that allows distinguishing between different primitive or constraint types and their parameters, while encouraging our model to re-use information across related parameters, encoding shared structure. A particular highlight of our work is the ability to produce primitives linked via constraints that enables the final output to be further regularized via a constraint solver. We evaluate our model by demonstrating constraint prediction for given sets of primitives and full sketch generation from scratch, showing that our approach significantly out performs the state-of-the-art in CAD sketch generation.
dc.description.sponsorshipThis work was suppported in part by ARL grant W911NF2120104, and a Vannevar Bush Faculty Fellowship. We would like to acknowledge gifts from Adobe, Autodesk and the UCL AI Centre. We thank the KAUST Supercomputing Lab (KSL) for providing compute infrastructure. Finally, we thank the anonymous reviewers for their constructive comments.
dc.publisherNeural information processing systems foundation
dc.rightsArchived with thanks to Neural information processing systems foundation
dc.titleSketchGen: Generating Constrained CAD Sketches
dc.typeConference Paper
dc.contributor.departmentComputer Science Program
dc.contributor.departmentVisual Computing Center (VCC)
dc.contributor.departmentComputer, Electrical and Mathematical Science and Engineering (CEMSE) Division
dc.conference.date2021-12-06 to 2021-12-14
dc.conference.name35th Conference on Neural Information Processing Systems, NeurIPS 2021
dc.conference.locationVirtual, Online
dc.contributor.institutionAdobe Research
dc.contributor.institutionUniversity of Leeds, United Kingdom
dc.contributor.institutionUniversity College London, United Kingdom
dc.contributor.institutionStanford University, United States
kaust.personPara, Wamiq Reyaz
kaust.personBhat, Shariq Farooq
kaust.personWonka, Peter
kaust.acknowledged.supportUnitSupercomputing Lab (KSL)

Files in this item

Accepted manuscript

This item appears in the following Collection(s)

Show simple item record


*Selected version