Gaussian-Product Subdivision Surfaces

Reinhold Preiner, Tamy Boubekeur, Michael Wimmer

Research output: Contribution to conferencePaperResearchpeer-review

Abstract

Probabilistic distribution models like Gaussian mixtures have shown great potential for improving both the quality and speed of several geometric operators. This is largely due to their ability to model large fuzzy data using only a reduced set of atomic distributions, allowing for large compression rates at minimal information loss. We introduce a new surface model that utilizes these qualities of Gaussian mixtures for the definition and control of a parametric smooth surface. Our approach is based on an enriched mesh data structure, which describes the probability distribution of spatial surface locations around each vertex via a Gaussian covariance matrix. By incorporating this additional covariance information, we show how to define a smooth surface via a non-linear probabilistic subdivision operator based on products of Gaussians, which is able to capture rich details at fixed control mesh resolution. This entails new applications in surface reconstruction, modeling, and geometric compression.
Original languageEnglish
Number of pages11
Publication statusPublished - Jul 2019
EventSIGGRAPH 2019 - Los Angeles, United States
Duration: 28 Jul 20191 Aug 2019

Conference

ConferenceSIGGRAPH 2019
CountryUnited States
Period28/07/191/08/19

Fingerprint

Surface reconstruction
Covariance matrix
Probability distributions
Data structures

Keywords

  • Gaussian Mixture
  • GMM
  • Subdivision Surface
  • Gaussian Product
  • Covariance Mesh
  • Nonlinear Subdivision

Cite this

Preiner, R., Boubekeur, T., & Wimmer, M. (2019). Gaussian-Product Subdivision Surfaces. Paper presented at SIGGRAPH 2019, United States.

Gaussian-Product Subdivision Surfaces. / Preiner, Reinhold; Boubekeur, Tamy; Wimmer, Michael.

2019. Paper presented at SIGGRAPH 2019, United States.

Research output: Contribution to conferencePaperResearchpeer-review

Preiner, R, Boubekeur, T & Wimmer, M 2019, 'Gaussian-Product Subdivision Surfaces' Paper presented at SIGGRAPH 2019, United States, 28/07/19 - 1/08/19, .
Preiner R, Boubekeur T, Wimmer M. Gaussian-Product Subdivision Surfaces. 2019. Paper presented at SIGGRAPH 2019, United States.
Preiner, Reinhold ; Boubekeur, Tamy ; Wimmer, Michael. / Gaussian-Product Subdivision Surfaces. Paper presented at SIGGRAPH 2019, United States.11 p.
@conference{b78e94803ee245b7b41cb502f3004739,
title = "Gaussian-Product Subdivision Surfaces",
abstract = "Probabilistic distribution models like Gaussian mixtures have shown great potential for improving both the quality and speed of several geometric operators. This is largely due to their ability to model large fuzzy data using only a reduced set of atomic distributions, allowing for large compression rates at minimal information loss. We introduce a new surface model that utilizes these qualities of Gaussian mixtures for the definition and control of a parametric smooth surface. Our approach is based on an enriched mesh data structure, which describes the probability distribution of spatial surface locations around each vertex via a Gaussian covariance matrix. By incorporating this additional covariance information, we show how to define a smooth surface via a non-linear probabilistic subdivision operator based on products of Gaussians, which is able to capture rich details at fixed control mesh resolution. This entails new applications in surface reconstruction, modeling, and geometric compression.",
keywords = "Gaussian Mixture, GMM, Subdivision Surface, Gaussian Product, Covariance Mesh, Nonlinear Subdivision",
author = "Reinhold Preiner and Tamy Boubekeur and Michael Wimmer",
year = "2019",
month = "7",
language = "English",
note = "SIGGRAPH 2019 ; Conference date: 28-07-2019 Through 01-08-2019",

}

TY - CONF

T1 - Gaussian-Product Subdivision Surfaces

AU - Preiner, Reinhold

AU - Boubekeur, Tamy

AU - Wimmer, Michael

PY - 2019/7

Y1 - 2019/7

N2 - Probabilistic distribution models like Gaussian mixtures have shown great potential for improving both the quality and speed of several geometric operators. This is largely due to their ability to model large fuzzy data using only a reduced set of atomic distributions, allowing for large compression rates at minimal information loss. We introduce a new surface model that utilizes these qualities of Gaussian mixtures for the definition and control of a parametric smooth surface. Our approach is based on an enriched mesh data structure, which describes the probability distribution of spatial surface locations around each vertex via a Gaussian covariance matrix. By incorporating this additional covariance information, we show how to define a smooth surface via a non-linear probabilistic subdivision operator based on products of Gaussians, which is able to capture rich details at fixed control mesh resolution. This entails new applications in surface reconstruction, modeling, and geometric compression.

AB - Probabilistic distribution models like Gaussian mixtures have shown great potential for improving both the quality and speed of several geometric operators. This is largely due to their ability to model large fuzzy data using only a reduced set of atomic distributions, allowing for large compression rates at minimal information loss. We introduce a new surface model that utilizes these qualities of Gaussian mixtures for the definition and control of a parametric smooth surface. Our approach is based on an enriched mesh data structure, which describes the probability distribution of spatial surface locations around each vertex via a Gaussian covariance matrix. By incorporating this additional covariance information, we show how to define a smooth surface via a non-linear probabilistic subdivision operator based on products of Gaussians, which is able to capture rich details at fixed control mesh resolution. This entails new applications in surface reconstruction, modeling, and geometric compression.

KW - Gaussian Mixture

KW - GMM

KW - Subdivision Surface

KW - Gaussian Product

KW - Covariance Mesh

KW - Nonlinear Subdivision

M3 - Paper

ER -