Interactive design of probability density functions for shape grammars

Abstract

KAUST Mark Pauly EPFLFigure 1: (Left) Random models generated from a probabilistic building grammar. Although these models are visually plausible, they do not comply with a design scenario which also requires architectural plausibility, i.e. matching styles of ground floors, upper floors, and roofs (B1, see Table 2). (Right) Our framework takes user specified preference scores as input and learns a new model probability density function (pdf) which samples models (with consistent style) proportionally to their predicted preference scores. In this design scenario, office buildings received a higher preference score. AbstractA shape grammar defines a procedural shape space containing a variety of models of the same class, e.g. buildings, trees, furniture, airplanes, bikes, etc. We present a framework that enables a user to interactively design a probability density function (pdf) over such a shape space and to sample models according to the designed pdf. First, we propose a user interface that enables a user to quickly provide preference scores for selected shapes and suggest sampling strategies to decide which models to present to the user to evaluate. Second, we propose a novel kernel function to encode the similarity between two procedural models. Third, we propose a framework to interpolate user preference scores by combining multiple techniques: function factorization, Gaussian process regression, autorelevance detection, and l 1 regularization. Fourth, we modify the original grammars to generate models with a pdf proportional to the user preference scores. Finally, we provide evaluations of our user interface and framework parameters and a comparison to other exploratory modeling techniques using modeling tasks in five example shape spaces: furniture, low-rise buildings, skyscrapers, airplanes, and vegetation. *

RoboGrammar

“…Lipp et al [2008] develop a framework for interactive grammar editing. Dang et al [2015] expand shape grammars with probability density functions defined through an interactive design exploration tool to obtain designs with higher preference score. A probabilistic grammar is also used in [Liu et al 2014] to parse unseen scenes and assign segmentation, labels, and object hierarchies.…”

A 3D shape generative method for aesthetic product design

“…Therefore, some works use different algorithms to ensure a homogeneous space exploration Khan & Awan, 2018;Khan & Gunpinar, 2018). (Dang et al, 2015) propose a probability density function combined with machine learning to avoid pure random concepts and focus generation in desired areas of the design space.…”

“…In this case, the two solutions highlighted in column B present the same traits (array and non-uniform scale) despite coming from different configurations. This mechanism emulates the "probability density function" used in (Dang et al, 2015). Assigning higher probabilities to desired traits, the designer may control the style of the product while keeping other parameters variable.…”

Interactive 3D Modeling with a Generative Adversarial Network

“…For example, [17], [1], and others describe methods for interpolation between surfaces with consistent parameterizations. More recently, probabilistic models of part hierarchies [16], [14] and grammars of shape features [8] have been learned from collections and used to assist synthesis of new shapes. However, these methods rely on specific hand-selected models and thus are not general to all types of shapes.…”