IEEE VIS 2024 Content: FunM^2C: A Filter for Uncertainty Visualization of Multivariate Data on Multi-Core Devices

FunM^2C: A Filter for Uncertainty Visualization of Multivariate Data on Multi-Core Devices

Gautam Hari - Indiana University Bloomington, Bloomington, United States

Nrushad A Joshi - Indiana University Bloomington, Bloomington, United States

Zhe Wang - Oak Ridge National Laboratory, Oak Ridge, United States

Qian Gong - Oak Ridge National Laboratory, Oak Ridge, United States

David Pugmire - Oak Ridge National Laboratory, Oak Ridge, United States

Kenneth Moreland - Oak Ridge National Laboratory, Oak Ridge, United States

Chris R. Johnson - University of Utah, Salt Lake City, United States

Scott Klasky - Oak Ridge National Laboratory, Oak Ridge, United States

Norbert Podhorszki - Oak Ridge National Laboratory, Oak Ridge, United States

Tushar M. Athawale - Oak Ridge National Laboratory, Oak Ridge, United States

Room: Bayshore VI

2024-10-14T12:30:00ZGMT-0600Change your timezone on the schedule page
2024-10-14T12:30:00Z
Exemplar figure, described by caption below
A simulation of the Deep Water Impact. From Left to Right, the images are a) Original Dataset, b) Compressed data without uncertainty, and c) Compressed data with uncertainty. The colors of the Uncertainty image range from transparent deep purple regions that indicate positions of lower probability, whereas the less transparent bright yellow regions indicate positions of higher probability. Uncertainty visualization recovers key topological structures, such as the rib-like formations (e.g., rib-like structure in the inset views), which appear broken in traditional mean-field visualization. This probabilistic approach of uncertainty visualization allows for the recovery of potentially important features in uncertain data.
Fast forward
Abstract

Uncertainty visualization is an emerging research topic in data vi- sualization because neglecting uncertainty in visualization can lead to inaccurate assessments. In this short paper, we study the prop- agation of multivariate data uncertainty in visualization. Although there have been a few advancements in probabilistic uncertainty vi- sualization of multivariate data, three critical challenges remain to be addressed. First, state-of-the-art probabilistic uncertainty visual- ization framework is limited to bivariate data (two variables). Sec- ond, the existing uncertainty visualization algorithms use compu- tationally intensive techniques and lack support for cross-platform portability. Third, as a consequence of the computational expense, integration into interactive production visualization tools is imprac- tical. In this work, we address all three issues and make a threefold contribution. First, we generalize the state-of-the-art probabilis- tic framework for bivariate data to multivariate data with a arbi- trary number of variables. Second, through utilization of VTK-m’s shared-memory parallelism and cross-platform compatibility fea- tures, we demonstrate acceleration of multivariate uncertainty visu- alization on different many-core architectures, including OpenMP and AMD GPUs. Third, we demonstrate the integration of our al- gorithms with the ParaView software. We demonstrate utility of our algorithms through experiments on multivariate simulation data.