Saved in:
Bibliographic Details
Main Authors: Dax, Maximilian, Berbel, Jordi, Stria, Jan, Guibas, Leonidas, Bergmann, Urs
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2501.17044
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909469480321024
author Dax, Maximilian
Berbel, Jordi
Stria, Jan
Guibas, Leonidas
Bergmann, Urs
author_facet Dax, Maximilian
Berbel, Jordi
Stria, Jan
Guibas, Leonidas
Bergmann, Urs
contents We generate abstractions of buildings, reflecting the essential aspects of their geometry and structure, by learning to invert procedural models. We first build a dataset of abstract procedural building models paired with simulated point clouds and then learn the inverse mapping through a transformer. Given a point cloud, the trained transformer then infers the corresponding abstracted building in terms of a programmatic language description. This approach leverages expressive procedural models developed for gaming and animation, and thereby retains desirable properties such as efficient rendering of the inferred abstractions and strong priors for regularity and symmetry. Our approach achieves good reconstruction accuracy in terms of geometry and structure, as well as structurally consistent inpainting.
format Preprint
id arxiv_https___arxiv_org_abs_2501_17044
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Synthesizing 3D Abstractions by Inverting Procedural Buildings with Transformers
Dax, Maximilian
Berbel, Jordi
Stria, Jan
Guibas, Leonidas
Bergmann, Urs
Computer Vision and Pattern Recognition
Artificial Intelligence
Machine Learning
We generate abstractions of buildings, reflecting the essential aspects of their geometry and structure, by learning to invert procedural models. We first build a dataset of abstract procedural building models paired with simulated point clouds and then learn the inverse mapping through a transformer. Given a point cloud, the trained transformer then infers the corresponding abstracted building in terms of a programmatic language description. This approach leverages expressive procedural models developed for gaming and animation, and thereby retains desirable properties such as efficient rendering of the inferred abstractions and strong priors for regularity and symmetry. Our approach achieves good reconstruction accuracy in terms of geometry and structure, as well as structurally consistent inpainting.
title Synthesizing 3D Abstractions by Inverting Procedural Buildings with Transformers
topic Computer Vision and Pattern Recognition
Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2501.17044