Home

Awesome

Sketch2CAD: Sequential CAD Modeling by Sketching in Context

Introduction

This repository contains the implementation of Sktech2CAD proposed in our SIGGRAPH Asia 2020 paper.

It contains two parts: 1) network training, 2) training dataset and trained network deployment (e.g., for interactive modeling).

The code is released under the MIT license.

Network training

💡 Great News: we have released the docker image for network training, which will greatly reduce the configuration burden, please check the networkTraining folder for more details.

This part contains the Python code for building, training and testing the nueral network using TensorFlow.

Please read README file within the networkTraining folder for more details.

Training dataset and network deployment

This part contains the code for deploying the trained network in a C++ project that can be an interactive 3D modeling application. It also provides instructions to download the training dataset we generated, and our trained networks.

Please read the README file in dataAndModel folder for more details.

Prototype System

We released the prototype system for research use, feel free to download and try it. More instructions to download it, please refer to the project page.

Citation

If you use our code or model, please cite our paper:

@Article{Li:2020:Sketch2CAD, 
	Title = {Sketch2CAD: Sequential CAD Modeling by Sketching in Context}, 
    	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} 
}

Contact

Any question you could contact Changjian Li (chjili2011@gmail.com) or Hao Pan (haopan@microsoft.com) for help.