SkexGen: Autoregressive Generation of CAD Construction Sequences with Disentangled Codebooks

Xiang Xu, Karl D.D. Willis, Joseph G Lambourne, Chin-Yi Cheng, Pradeep Kumar Jayaraman, Yasutaka Furukawa
Proceedings of the 39th International Conference on Machine Learning, PMLR 162:24698-24724, 2022.

Abstract

We present SkexGen, a novel autoregressive generative model for computer-aided design (CAD) construction sequences containing sketch-and-extrude modeling operations. Our model utilizes distinct Transformer architectures to encode topological, geometric, and extrusion variations of construction sequences into disentangled codebooks. Autoregressive Transformer decoders generate CAD construction sequences sharing certain properties specified by the codebook vectors. Extensive experiments demonstrate that our disentangled codebook representation generates diverse and high-quality CAD models, enhances user control, and enables efficient exploration of the design space. The code is available at https://samxuxiang.github.io/skexgen.

Cite this Paper

BibTeX
@InProceedings{pmlr-v162-xu22k, title = {{S}kex{G}en: Autoregressive Generation of {CAD} Construction Sequences with Disentangled Codebooks}, author = {Xu, Xiang and Willis, Karl D.D. and Lambourne, Joseph G and Cheng, Chin-Yi and Jayaraman, Pradeep Kumar and Furukawa, Yasutaka}, booktitle = {Proceedings of the 39th International Conference on Machine Learning}, pages = {24698--24724}, year = {2022}, editor = {Chaudhuri, Kamalika and Jegelka, Stefanie and Song, Le and Szepesvari, Csaba and Niu, Gang and Sabato, Sivan}, volume = {162}, series = {Proceedings of Machine Learning Research}, month = {17--23 Jul}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v162/xu22k/xu22k.pdf}, url = {https://proceedings.mlr.press/v162/xu22k.html}, abstract = {We present SkexGen, a novel autoregressive generative model for computer-aided design (CAD) construction sequences containing sketch-and-extrude modeling operations. Our model utilizes distinct Transformer architectures to encode topological, geometric, and extrusion variations of construction sequences into disentangled codebooks. Autoregressive Transformer decoders generate CAD construction sequences sharing certain properties specified by the codebook vectors. Extensive experiments demonstrate that our disentangled codebook representation generates diverse and high-quality CAD models, enhances user control, and enables efficient exploration of the design space. The code is available at https://samxuxiang.github.io/skexgen.} }
Endnote
%0 Conference Paper %T SkexGen: Autoregressive Generation of CAD Construction Sequences with Disentangled Codebooks %A Xiang Xu %A Karl D.D. Willis %A Joseph G Lambourne %A Chin-Yi Cheng %A Pradeep Kumar Jayaraman %A Yasutaka Furukawa %B Proceedings of the 39th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2022 %E Kamalika Chaudhuri %E Stefanie Jegelka %E Le Song %E Csaba Szepesvari %E Gang Niu %E Sivan Sabato %F pmlr-v162-xu22k %I PMLR %P 24698--24724 %U https://proceedings.mlr.press/v162/xu22k.html %V 162 %X We present SkexGen, a novel autoregressive generative model for computer-aided design (CAD) construction sequences containing sketch-and-extrude modeling operations. Our model utilizes distinct Transformer architectures to encode topological, geometric, and extrusion variations of construction sequences into disentangled codebooks. Autoregressive Transformer decoders generate CAD construction sequences sharing certain properties specified by the codebook vectors. Extensive experiments demonstrate that our disentangled codebook representation generates diverse and high-quality CAD models, enhances user control, and enables efficient exploration of the design space. The code is available at https://samxuxiang.github.io/skexgen.
APA
Xu, X., Willis, K.D., Lambourne, J.G., Cheng, C., Jayaraman, P.K. & Furukawa, Y.. (2022). SkexGen: Autoregressive Generation of CAD Construction Sequences with Disentangled Codebooks. Proceedings of the 39th International Conference on Machine Learning, in Proceedings of Machine Learning Research 162:24698-24724 Available from https://proceedings.mlr.press/v162/xu22k.html.

Related Material