Seminar

The Telecom ParisTech Computer Graphics organizes scientific seminars on a regular basis. Please contact Tamy Boubekeur if you want to be added to the dedicated mailing list.

Next seminar

WHAT: Probabilistic Subdivision Surfaces using Covariance Meshes

WHO: Dr. Reinhold Preiner, TU Graz, https://www.researchgate.net/profile/Reinhold_Preiner

WHEN: Monday September, 17th, 3pm

WHERE: Room C46, Building C, Ground Floor, Telecom ParisTech, 46 rue Barrault, 75013 Paris (Metro line 5 “Place d’Italie”, line 6 “Corvisart”, line 7 “Tolbiac”)

Abstract

For the efficient processing of fuzzy geometric data like noisy point sets, probabilistic distribution models like Gaussian mixtures have shown great potential for improving both the quality and speed of tasks like registration, filtering or resampling. 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.
In this talk we introduce a new probabilistic surface model, that utilizes these qualities of Gaussian mixtures for the definition of a parametric probabilistic surface.
Our surface model 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 smooth surfaces via a non-linear probabilistic subdivision operator, which is able to capture rich details at fixed control mesh resolution.
This entails new future tools in surface reconstruction, modelling, and geometric compression.

Short Bio

Reinhold Preiner has been part of the Institute of Computer Graphics at TU Wien from 2010 to 2017, where he received his Master and PhD in Visual Computing. Since 2018 he is a Postdoctoral Reseracher at the Institute of Computer Graphics and Knowledge Visualization, TU Graz.
His research work and interests include efficient dynamic point-cloud processing, surface reconstruction and rendering, and probabilistic methods for geometry processing.