1University College London     2 Microsoft Research Asia
3Inria, Université Côte d’Azur     4 Adobe Research

Siggrpah Asia 2020

Industrial designers commonly decompose complex shapes into box-like primitives, which they refine by drawing cuts and roundings, or by adding and substracting smaller parts (a, © Koos Eissen and Roselien Steur). Users of Sketch2CAD follow similar sketching steps (b), which our system interprets as parametric modeling operations (c) to automatically output a precise, compact, and editable CAD model (d).

###### Presentation Video

Given an existing shape and input sketch strokes (shown in orange) for the current operation, we first obtain the maps of sketch and local context (i.e., depth and normal), which are fed to the operator classification and segmentation networks. The classified operator type, sweep in this example, is used to select the output base face and curve segmentation maps, based on which the parameters defining the operator are fitted, via an optimization, to recover the sketched operation instance. The recovered operator is then applied to the existing shape to produce the updated model; meanwhile, the operation is pushed into the protocol list.

We synthetically generated 10k protocols of diverse lengths for procedurally generating 40k training shapes. For each protocol, we execute it up to the last operation, for which the sketch curves are built and overlaid on the built shape. The sketch curves and existing shape are rendered in proper viewpoints to generate the input sketch and local context maps, as well as the ground truth face and curve segmentation maps, which are used to train the operator classifier and the corresponding segmentation network.

###### Results Gallery

Various modeling sequences created during design sessions using Sketch2CAD. The corresponding protocol steps are shown in the supplemental material. Please also refer to the supplementary video.

###### Bibtex
@Article{Li:2020:Sketch2CAD,
Author = {Changjian Li and Hao Pan and Adrien Bousseau and Niloy J. Mitra},
Journal = {ACM Trans. Graph. (Proceedings of SIGGRAPH Asia 2020)},
Year = {2020},
Number = {6},
Volume = {39},
Pages={164:1--164:14},
numpages = {14},
DOI={https://doi.org/10.1145/3414685.3417807},
Publisher = {ACM}
}

###### Acknowledgements

The authors would like to thank the reviewers for their valuable and detailed suggestions, the user evaluation participants and Nathan Carr, Yuxiao Guo, Zhiming Cui for the valuable discussions. The work of Niloy was supported by ERC Grant (SmartGeometry 335373), Google Faculty Award and gifts from Adobe, and the work of Adrien was supported by ERC Starting Grant D3 (ERC-2016-STG 714221), research and software donations from Adobe. Finally, Changjian Li wants to thank, in particular, the endless and invaluable love and supports from Huahua Guo over the tough time due to COVID-19.