Projects
VisTrails
VisTrails is a new visualization system whose goal is to provide scientists
with an infrastructure that helps the exploration process
inherent in developing effective visualizations. VisTrails
emphasizes the importance of process provenance: it
transparently keeps track of the visualizations as they are
created. This facilitates exploration and reproducibility - users
can always very easily go back to previous results.
We also show how to reuse provenance to
help users make better sense of their
tasks. VisTrails was originally created for scientific
visualization, but its infrastructure is equally well-suited for
more general computational tasks - in particular those involving
multiple separate libraries.
VisComplete: Automating Suggestions for Visualization
Pipelines (preprint).
IEEE Trans. Vis. Comp. Graph., 14(6):1691-1698, 2008 (Vis 2008).
Examining Statistics of Workflow Evolution Provenance: A
First Study. SSDBM 2008.
Querying and Creating Visualizations by
Analogy. IEEE Trans. Vis. Comp. Graph. 13(6):1560-1567 (Vis 2007). Best paper award.
Tackling the Provenance Challenge One Layer at a
Time. Concurrency Computat.:
Pract. Exper. 20(5):473-483, 2008.
Managing Rapidly Evolving Scientific Workflows. IPAW 2006.
VisTrails: Visualization Meets Data Management. SIGMOD 2006.
Managing the Evolution of Dataflows with
VisTrails. SciFlow 2006.
VisTrails: Enabling Interactive Multiple-View
Visualizations. Vis 2005.
Geometry Processing
One problem I have explored in geometry processing is that of
generating high-quality triangle meshes from other
representations. I have helped develop an advancing front method that is applicable to a wide range of
representations. In recent collaborations, we have proposed variants of
Marching Cubes that generate triangle meshes with good
quality. We have also recently investigated the relation
between volume integrals and isosurface integrals. In
particular, our (very simple) use of the co-area formula gives a clean
definition of the continuous analogue of a histogram, and direct
extensions of the popular contour spectrum functions.
Revisiting Histograms and Isosurface
Statistics (preprint).
IEEE Trans. Vis. Comp. Graph., 14(6):1659-1666, 2008. (Vis 2008)
More here.
Edge Groups: An Approach For
Understanding The Quality of Marching Cubes (preprint).
IEEE Trans. Vis. Comp. Graph., 14(6):1651-1658, 2008. (Vis 2008)
More here.
Edge Transformations for Improving Mesh Quality of Marching
Cubes. IEEE Trans. Vis. Comp. Graph.,
15(1):150-159, 2009.
High-Quality Extraction of Isosurfaces from Regular and
Irregular Grids. IEEE
Trans. Vis. Comp. Graph. 12(5):1205-1212. (Vis 2006).
Direct (Re)Meshing for Efficient Surface
Processing. Comp. Graph. Forum, 25(3):527-536. (EG 2006).
Triangulating Point-Set Surfaces with Bounded
Error. SGP 2005.
Point Set Surfaces
Bandwidth Selection and Reconstruction Quality in Point-Based
Surfaces. IEEE Trans. Vis. Comp. Graph., to appear, 2009.
Optimal Bandwidth Selection for MLS Surfaces. SMI
2008. Best paper award.
A Unified Projection Operator for MLS
Surfaces. Tech report.
|
|
|
Other
Marching Cubes
with no Skinny Triangles. IEEE Computing in Science and
Engineering, (to appear)
Direct Volume Rendering: A 3D Plotting Technique for
Scientific Data. IEEE Computing in Science and
Engineering, 10(1):88-92, 2008.
Hardware-Assisted
Point-Based Volume Rendering of Tetrahedral
Meshes. SIBGRAPI 2007.
Practical CFD
Simulations on the GPU using
SMAC. Comp. Graph. Forum, 24(4):715-728, 2005. source
