VECNA: Visual Exploration, Comparison and Analysis of Reconstructed Spatiotemporal Scientific Simulation Data
Year: 2025
Type: conference-paper
Abstract: Data-driven sampling and reconstruction techniques are increasingly being employed in scientific computing applications to aggressively reduce data volumes while retaining the crucial features of spatiotemporal datasets. Such data must be reconstructed for analysis, but it is difficult for domain experts to assess reconstruction quality, particularly given the pace at which new methods are being developed and a lack of support in existing tools. To help address this, we introduce VECNA, a visual analytics system for exploring and comparing reconstructed scientific datasets. Developed through collaboration with high-performance computing researchers, VECNA enables intuitive qualitative and quantitative comparisons among diverse reconstruction methodologies. We validate VECNA via a usage scenario and empirical expert assessments to demonstrate its efficacy in empowering users to discern nuances in reconstruction quality and identify regions of interest within datasets, facilitating more informed subsequent analyses.
Source: Eurographics
DOI: 10.2312/evs.20251082
Authors: Mishra, Aditi, Biswas, Ayan, Bryan, Chris
Institutions: Los Alamos National Laboratory, Arizona State University
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W6888843259
FWCI:
Cites:
Cited by:
Related to: