Proceedings: GI 2019

Controlling Procedural Modelling Interactively with Guiding Curves

Dave Pagurek van Mossel (University of Waterloo), Abhishek Madan (University of Waterloo), Tai Meng Liu (University of Waterloo), Paul Bardea (University of Waterloo), Andrew McBurney (University of Waterloo)

Proceedings of Graphics Interface 2019: Kingston, Ontario, 28 - 31 May 2019

  • BibTex

    @inproceedings{Pagurek van Mossel:2019:10.20380/GI2019.12,
    author = {Pagurek van Mossel, Dave and Madan, Abhishek and Liu, Tai Meng and Bardea, Paul and McBurney, Andrew},
    title = {Controlling Procedural Modelling Interactively with Guiding Curves},
    booktitle = {Proceedings of Graphics Interface 2019},
    series = {GI 2019},
    year = {2019},
    issn = {0713-5424},
    isbn = {978-0-9947868-4-5},
    location = {Kingston, Ontario},
    numpages = {8},
    doi = {10.20380/GI2019.12},
    publisher = {Canadian Information Processing Society},
  • Supplementary Media


Grammar-based procedural modelling on its own produces a larger space of generated models than is artistically desirable. Probabilistic sampling techniques can help search this result space for models that best fit a set of constraints. We aim to provide a useful probabilistic search function that can be run at interactive rates to enable the short feedback loops artists require for incremental, exploratory design. We present a constraint for use in Sequential Monte Carlo optimization where artists draw curves to guide the generation of models. The high-level structure of models can be intuitively specified by our constraint framework, allowing for variation in low-level details to be automatically filled in. We present a real-time model editor to demonstrate the artistic utility of our method.