Fast 3D Indoor Scene Synthesis with Discrete and Exact Layout Pattern Extraction

Hausdorff distance Disjoint sets
DOI: 10.48550/arxiv.2002.00328 Publication Date: 2020-01-01
ABSTRACT
We present a fast framework for indoor scene synthesis, given room geometry and list of objects with learnt priors. Unlike existing data-driven solutions, which often extract priors by co-occurrence analysis statistical model fitting, our method measures the strengths spatial relations tests complete randomness (CSR), extracts complex based on samples ability to accurately represent discrete layout patterns. With extracted priors, achieves both acceleration plausibility partitioning input into disjoint groups, followed optimization Hausdorff metric. Extensive experiments show that is capable measuring more reasonable among simultaneously generating varied arrangements in seconds.
SUPPLEMENTAL MATERIAL
Coming soon ....
REFERENCES ()
CITATIONS ()
EXTERNAL LINKS
PlumX Metrics
RECOMMENDATIONS
FAIR ASSESSMENT
Coming soon ....
JUPYTER LAB
Coming soon ....