- Amy Hawkins - Camera control and surface modeling
- Camera Interpolation: Amy applied Linear matrix interpolation to camera interpolation. This involved re-writing the projection matrix into a more suitable form, looking at three types of interpolation (piece-wise linear, smooth approximating and smooth interpolating) and altering the interpolation when the rotation is too big.
Surface parameterization: There are several theoretical papers out there for cutting n-holed tori open along their generator lines (most using a version of minimum spanning tree). Actually implementing these algorithms on real meshes, then using them to create a parameterization of the hyperbolic disk is a challenging problem.
- Ross Sowell - Interactive extraction and editing of surfaces from 3D imaging data.
Ross has taken a faculty position at Cornell College, 2011
For his PhD, Ross developed an interface for creating, editing, and reviewing surfaces made from oblique contours. Specifically, he looked at how accurate physicians are at drawing these contours, how best to navigate through 3D images to place the contours, and what are the best contours to draw.
Joint work with: Christopher Abrams and Daniel Low (Radiology at UCLA), Lu Liu and Tao Ju (Computer Science).
Ross was an Imaging Sciences Pathway participant, 2007.
- Nisha Sudarsanam - Non-linear projections and camera control
- Nisha started off looking at non-linear projection systems, and quickly discovered that defining one camera was difficult, much less multiple ones. This led to the CubeWidget, a screen-based interactive widget for camera control. The CubeWidget both simplifies the IBar and provides better feedback for novice users. She is now looking at screen-space widgets for specific non-linear projections. The first of these supports curvi-linear perspective, where vanishing lines form sinusoids instead of straight lines.
- Reynold Bailey - Perception-based image editing and subtle gaze direction
- Rennie worked on in-depth studies of how we perceive color and luminance. He used these to create image-editing techniques for creating more intuitive, perception-based changes of perceived depth and motion. He completed his PhD in August of 2007.
Rennie is now a professor at Rochester Institute of Technology where he received his CAREER award in 2009. We are continuing to work on subtle gaze direction techniques, which is joint work with Ann McNamara at Texas A&M.
- Timothy D. Gatzke - Surface comparison
- Timothy developed metrics for comparing and describing local surface shape. He extended this work to surface feature identification (is this shape a peak? a ridge?) and mutually consistent parameterizations of similar, but not identical, shapes. He worked full time at Boeing (!) while completing his PhD (completed in 2006). Sadly, Tim passed away in March of 2011. He will be missed.
- Rob Glaubius - Surface fitting using manifolds
- Rob wass actually Bill Smart's student working on reinforcement learning for robots using manifolds, and for process control (joint work with Chris Gill ). Fitting value functions to sample data is a (much harder) version of surface fitting. Rob applied chart placement approaches to the manifold surface fitting problem.
- Mark Schroering - 3D input device
- Mark developed a 3D sketching device based on a (tracked) piece of paper and a camera.
- Raquel Bujans - Mesh editing
- Raquel applied both widget-based and sketch-based techniques to the problem of mesh editing.
- Bill Niebridge - GPU techniques
- Bill took a continuous, overlapping spherical parameterization technique and ported it to the GPU as an alternative to cube mapping. Since the mapping is continuous and overlapping, it is an ideal representation for capturing and storing environment maps.
- Andrew Schoewe - GPU Genetic textures
- Andrew implemented genetic textures in the GPU.
- Martin Hassett - Implicit surface editing
- Martin is now at Boeing.
- Nicholas Haddad - Sound
- Nick applied texture synthesis approaches to sound.
- Ying Huang - Texture synthesis
- Ying looked at data capture for view-dependent texture synthesis.
- Zachary Byers - Lewis the robotic photographer
- Zach was one of the primary architects of the Lewis robotic photographer system. He is now married and working at Southwestern Bell.
- Michael Dixon - Lewis the robotic photographer
- Michael was the other architect of the Lewis project. He worked in our lab for several years and is now at Willow Garage, as of 2011.
- Christopher Kulla - Texturing
- Chris looked at ways to generate texture variation to represent shading. His primary love is rendering; he is the Source forge lead on Sunflow, spent several years as the research department for Reel Fx, and is now at Sony Image works
- Tim Blakeley - Finger-tip UI
- Tim is experimenting with a 3D tracker-based system for 3D user-interfaces. This is a very low-end system that tracks just the finger tips. Tim also has done a lot of GPU implementation, including the non-linear projection code.
- Ly Phan - Consistent parameterization of developing or similar surfaces.
Ly graduated in 2012.
Ly's PhD was in the area of surface parameterization and comparison for biological applications. Given two surfaces that are similar, but not exactly the same, how do you create a correspondence between them? Establishing this correspondence is a first step towards making quantitative comparisons between shapes. Our basic approach is to use stress/strain energy minimization, along with feature landmarks, to build this correspondence. This is joint work with Phil Bayly and his student Andy Knutsen, who graduated in 2009.
Previously, Ly worked on sketching reaction diffusion textures
Ly was partly funded through the Imaging Science Pathway.
- Nathan Dudley - Abstraction and realism
- Artists often depict texturally complicated objects using a mix of abstraction and realism. For example, the basic tree shape is outlined in shaded blobs, then detail added in where necessary. Nathan has adopted this approach for rendering trees generated by L-Studio.
- Leon Barrett - Non-linear cameras
- Leon is the guy who developed the four-point camera model, which has now been used for both camera interpolation and non-linear projection. He is now a graduate student at UC Berkeley
- Timothy Lin - 3D Painting
- Timothy wrote a first draft of a 3D painting system that models the complex interactions of physical paint (with depth).
- Mamta Datwani - Subtle gaze direction for video
- Extending the static gaze direction to video, taking into account motion, color, and illumination in the video.
- Hyun-Taek Kwon - Color editing
- How to edit color blends so the artist has a lot more control over how the blend happens.
- Kylia Miskell - Color editing
- How to edit color blends. She currently works for Robert Pless as a PhD student
- Ruosi Li - Building surfaces from point sets.
- Ruosi graduated as a master's student working on building a surface from a point set. This is a continuation of the work started by Ly Phan. She created a blended vector field and converting the current marching-cubes style algorithm to an oct tree. (Note: Jointly advised by Tao Ju). This work is available as part of ROS, the Robot Operating System from Willow Garage.