A5080292005- Data Visualization and Analytics
- Computer Graphics and Visualization Techniques
- Topological and Geometric Data Analysis
- Simulation Techniques and Applications
- Scientific Computing and Data Management
- Distributed and Parallel Computing Systems
- Embedded Systems Design Techniques
- Advanced Numerical Analysis Techniques
- Data Analysis with R
- Computational Geometry and Mesh Generation
- Probabilistic and Robust Engineering Design
- Image Retrieval and Classification Techniques
- Advanced Measurement and Metrology Techniques
- Quantum Computing Algorithms and Architecture
- Neurological disorders and treatments
- Parkinson's Disease Mechanisms and Treatments
- Advanced Multi-Objective Optimization Algorithms
- Cardiac electrophysiology and arrhythmias
- Advanced Vision and Imaging
- Quantum Information and Cryptography
- Manufacturing Process and Optimization
- Cell Image Analysis Techniques
- Systems Engineering Methodologies and Applications
- Time Series Analysis and Forecasting
- Seismic Imaging and Inversion Techniques
Oak Ridge National Laboratory
2021-2025
University of Utah
2018-2024
Los Alamos National Laboratory
2020
University of Florida
2013-2015
The problem of isosurface extraction in uncertain data is an important research and may be approached two ways. One can extract statistics (e.g., mean) from points visualize the extracted field. Alternatively, uncertainty, characterized by probability distributions, propagated through process. We analyze impact uncertainty on topology geometry algorithms. A novel, edge-crossing based approach proposed to predict underlying for data. derive a probabilistic version midpoint decider that...
We present a study of linear interpolation when applied to uncertain data. Linear is key step for isosurface extraction algorithms, and the uncertainties in data lead non-linear variations geometry extracted isosurface. an approach deriving probability density function random variable modeling positional uncertainty extraction. When quantified by uniform distribution, our provides closed-form characterization mentioned variable. This allows us derive, closed form, expected value as well...
Abstract In this paper, we study uncertainty quantification and visualization of orientation distribution functions (ODF), which corresponds to the diffusion profile high angular resolution imaging (HARDI) data. The shape inclusion probability (SIP) function is state‐of‐the‐art method for capturing ODF ensembles. current computing SIP with a volumetric basis exhibits computational memory costs, can be bottleneck integrating into HARDI techniques tools. We propose novel spherical sampling...
Morse complexes are gradient-based topological descriptors with close connections to theory. They widely applicable in scientific visualization as they serve important abstractions for gaining insights into the topology of scalar fields. Data uncertainty inherent fields due randomness their acquisition and processing, however, limits our understanding structural abstractions. We, therefore, explore an ensemble 2D that arises from coupled data uncertainty. We propose several statistical...
We present a nonparametric statistical framework for the quantification, analysis, and propagation of data uncertainty in direct volume rendering (DVR). The state-of-the-art DVR allows preserving transfer function (TF) ground truth when visualizing uncertain data; however, existing is restricted to parametric models uncertainty. In this paper, we address limitations by extending distributions. exploit quantile interpolation technique derive probability distributions representing viewing-ray...
We present a framework for the analysis of uncertainty in isocontour extraction. The marching squares (MS) algorithm reconstruction generates linear topology that is consistent with hyperbolic curves piecewise bilinear interpolation. saddle points interpolant cause topological ambiguity midpoint decider and asymptotic are well-known mathematical techniques resolving ambiguities. latter technique investigates data values at cell resolution. data, however, leads to underlying interpolation...
Visualization and analysis of multivariate data their uncertainty are top research challenges in visualization. Constructing fiber surfaces is a popular technique for visualization that generalizes the idea level-set univariate to data. In this paper, we present statistical framework quantify positional probabilities fibers extracted from uncertain bivariate fields. Specifically, extend state-of-the-art Gaussian models other parametric distributions (e.g., uniform Epanechnikov) more general...
Deep brain stimulation (DBS) is an established therapy for treating patients with movement disorders such as Parkinson's disease. Patient-specific computational modelling and visualisation have been shown to play a key role in surgical therapeutic decisions DBS. The models use imaging, magnetic resonance (MR) computed tomography (CT), determine the DBS electrode positions within patient's head. finite resolution of however, introduces uncertainty positions. settings optimal patient response...
Marching squares (MS) and marching cubes (MC) are widely used algorithms for level-set visualization of scientific data. In this paper, we address the challenge uncertainty topology cases MS MC uncertain scalar field data sampled on a uniform grid. The is challenging due to their exponential nature possibility multiple per cell We propose case count entropy-based techniques quantifying in when noise modeled with probability distributions. demonstrate applicability our independent correlated...
Visualizing the uncertainty of ensemble simulations is challenging due to large size and multivariate temporal features en-semble data sets. One popular approach studying ensembles analyzing positional level Probabilistic marching cubes a technique that performs Monte Carlo sampling Gaussian noise distributions for visualization However, suffers from high computational time, making interactive analysis impossible achieve. This paper introduces deep-learning-based learning level-set...
The efficiency of solar panels depends on the operating temperature. As panel temperature rises, drops. Thus, energy community aims to understand factors that influence temperature, which include wind speed, direction, turbulence, ambient mounting configuration, and cell material. We use high-resolution numerical simulations model flow thermal behavior idealized farms. Because these such complex behavior, advanced visualization techniques are needed investigate results. Here, we present 3D...
Color maps are a commonly used visualization technique in which data mapped to optical properties, e.g., color or opacity. maps, however, do not explicitly convey structures (e.g., positions and scale of features) within data. Topology-based visualizations reveal communicate underlying Although our understanding what types features captured by topological is good, people's perception those not. This paper evaluates the sensitivity topology-based isocontour, Reeb graph, persistence diagram...
Multi-resolution methods such as Adaptive Mesh Refinement (AMR) can enhance storage efficiency for HPC applications generating vast volumes of data. However, their applicability is limited and cannot be universally deployed across all applications. Furthermore, integrating lossy compression with multi-resolution techniques to further boost encounters significant barriers. To this end, we introduce an innovative workflow that facilitates high-quality data both uniform AMR simulations....
The VTK-m software library enables scientific visualization on exascale-class supercomputers. Exascale machines are particularly challenging for development in part because they use GPU accelerators to provide the vast majority of their computational throughput. Algorithmic designs GPUs and GPU-centric computing often deviate from those that worked well previous generations high-performance computers relied traditional CPUs. Fortunately, provides algorithms other accelerators. also a...
The increasing adoption of Deep Neural Networks (DNNs) has led to their application in many challenging scientific visualization tasks. While advanced DNNs offer impressive generalization capabilities, understanding factors such as model prediction quality, robustness, and uncertainty is crucial. These insights can enable domain scientists make informed decisions about data. However, inherently lack ability estimate uncertainty, necessitating new research construct robust uncertainty-aware...
This paper presents a novel end-to-end framework for closed-form computation and visualization of critical point uncertainty in 2D uncertain scalar fields. Critical points are fundamental topological descriptors used the analysis The inherent data (e.g., observational experimental data, approximations simulations, compression), however, creates regarding positions. Uncertainty positions, therefore, cannot be ignored, given their impact on downstream tasks. In this work, we study as function...
This paper presents a novel end-to-end framework for closed-form computation and visualization of critical point uncertainty in 2D uncertain scalar fields. Critical points are fundamental topological descriptors used the analysis The inherent data (e.g., observational experimental data, approximations simulations, compression), however, creates regarding positions. Uncertainty positions, therefore, cannot be ignored, given their impact on downstream tasks. In this work, we study as function...
Isosurface visualization is fundamental for exploring and analyzing 3D volumetric data. Marching cubes (MC) algorithms with linear interpolation are commonly used isosurface extraction visualization. Although easy to implement, it has limitations when the underlying data complex high-order, which case most real-world Linear can output vertices at wrong location. Its inability deal sharp features smaller than grid cells lead an incorrect holes broken pieces. Despite these limitations,...
We present a simple comparative framework for testing and developing uncertainty modeling in uncertain marching cubes implementations. The selection of model to represent the probability distribution values directly influences memory use, run time, accuracy an visualization algorithm. use entropy calculation on ensemble data establish expected result then compare from various models, including uniform, Gaussian, histogram, quantile models. Our results verify that models matching indeed match...