Fast 3D Indoor Scene Synthesis by Learning Spatial Relation Priors of Objects

Songhai Zhang, Shao-Kui Zhang, Wei-Yu Xie, Cheng-Yang Luo, Yongliang Yang, Hongbo Fu

Research output: Contribution to journalArticlepeer-review

181 Downloads (Pure)

Abstract

We present a framework for fast synthesizing indoor scenes, given a room geometry and a list of objects with learnt priors.Unlike existing data-driven solutions, which often learn priors by co-occurrence analysis and statistical model fitting, our methodmeasures the strengths of spatial relations by tests for complete spatial randomness (CSR), and learns discrete priors based onsamples with the ability to accurately represent exact layout patterns. With the learnt priors, our method achieves both acceleration andplausibility by partitioning the input objects into disjoint groups, followed by layout optimization using position-based dynamics (PBD)based on the Hausdorff metric. Experiments show that our framework is capable of measuring more reasonable relations amongobjects and simultaneously generating varied arrangements in seconds compared with the state-of-the-art works.

Original languageEnglish
Pages (from-to)1-1
Number of pages1
JournalIEEE Transactions on Visualization and Computer Graphics
Early online date12 Jan 2021
DOIs
Publication statusE-pub ahead of print - 12 Jan 2021

Keywords

  • 3D Indoor Scene Synthesis
  • Complete Spatial Randomness
  • Computational modeling
  • Furniture Objects Arrangement
  • Layout
  • Mathematical model
  • Optimization
  • Semantics
  • Task analysis
  • Three-dimensional displays

ASJC Scopus subject areas

  • Software
  • Signal Processing
  • Computer Vision and Pattern Recognition
  • Computer Graphics and Computer-Aided Design

Fingerprint

Dive into the research topics of 'Fast 3D Indoor Scene Synthesis by Learning Spatial Relation Priors of Objects'. Together they form a unique fingerprint.

Cite this