Intro

People

Courses

Sponsors
and
Friends

Research

Gallery of
Student
Projects

Papers

Video

Info
for Lab
Members

Some past graphics projects

This list contains some of the projects done within the graphics group over the past few years.

	Content-Preserving Warps for 3D Video Stabilization By: Feng Liu, Michael Gleicher, Hailin Jin and Aseem Agarwala We present a technique that transforms a video from a hand-held video camera so that it appears as if it were taken with a directed camera motion.
	Discovering Panoramas in Web Video By: Feng Liu, Yu-hen Hu and Michael Gleicher In this paper, we present a method to discover panoramic source images within widely available existing videos, such as videos on YouTube.
	Learning Color and Locality Cues for Moving Object Detection and Segmentation By: Feng Liu and Michael Gleicher This paper presents an algorithm for automatically detecting and segmenting a moving object from a monocular video.
	Re-Cinematography By: Michael Gleicher and Feng Liu Re-Cinematography is video stabilization taken to the next level: rather than just getting rid of some of the jitter, the methods try to figure out what camera movements might have been done by a professional with good equipment, and then alter the video to look like that.
	Molecular Surface Abstraction By: Greg Cipriano and Michael Gleicher In this paper we introduce a visualization technique that provides an abstracted view of the shape and spatio-physico-chemical properties of complex molecules.
	Video Retargeting: Automating Pan and Scan By: Feng Liu and Michael Gleicher In this paper, we introduce Video Retargeting that adapts video to better suit the target display, minimizing the important information lost.
	Visual Query Results from Protein Databases By: Aneesh Karve and Michael Gleicher Information visualization of PDB query results, focused on ligand binding and polymer structure. The visualization is designed to provide dense, scannable results in order to facilitate comparative analysis of numerous proteins and ligands.
	Parametric Motion Graphs By: Rachel Heck and Michael Gleicher We present an example-based motion synthesis technique that generates continuous streams of high-fidelity, controllable motion for interactive applications, such as video games.
	Virtual Videography By: Michael Wallick, Rachel Heck, Michael Gleicher The main idea behind Virtual Videography is to automatically edit video of some event that would be nice to record, but too costly or intrusive to place a camera crew, such as a classroom lecture.
	Splicing Upper-Body Actions with Locomotion By: Rachel Heck, Lucas Kovar, and Michael Gleicher We present a simple and efficient technique for synthesizing high-fidelity motions by attaching, or splicing,the upper-body action of one motion example to the lower-body locomotion of another.
	Automated Extraction and Parameterization of Motions in Large Data Sets By: Lucas Kovar and Michael Gleicher We show how to automatically extract sets of logically related motion segments from a database and then convert them into a continuous, intuitively parameterized space of actions.
	Physical Touchup of Human Motions By: Hyun Joon Shin, Lucas Kovar, Michael Gleicher We present an approach to increase physical plausibility of human motions that are edited while not explicit incorpating physics. This approach alters motions so that they approximately satisfy physical laws, such as zero moment point constraint, linear and angular momentum preservation.
	Skinning by Example By: Alex Mohr and Michael Gleicher We construct high-quality skins for characters that can run on traditional runtimes from a set of examples.
	Decimation Sensitive Decimation By: Alex Mohr and Michael Gleicher
	Registration Curves By: Lucas Kovar and Michael Gleicher Many motion editing algorithms are forms of a general operation called motion blending. We introduce registration curves, which allow a wider class of motions to be blended without manual input.
	Snap Together Motion By: Michael Gleicher, Hyun Joon Shin, Lucas Kovar, Andrew Jepsen We present an approach to character motion called Snap-Together Motion that addresses the unique demands of virtual environments. Snap-Together Motion (STM) preprocesses a corpus of motion capture examples into a set of short clips that can be concatenated to make continuous streams of motion.
	Skin Authoring By: Luke Tokheim A system for authoring linear blend skins. Provides a more standard painting interface plus our own direct manipulation method for setting vertex weights.
	Motion Graphs By: Lucas Kovar, Michael Gleicher, Fred Pighin We present a method for controlling motion capture data without sacrificing motion quality. Given a database of motion, we automatically construct seamless transitions to form a directed graph we call a motion graph. A user can then extract motions that meet a set of constraints. Our framework is applied to the particular problem of directing locomotion down arbitrary paths.
	Footskate Cleanup for Motion Capture Editing By: Lucas Kovar, John Schreiner, Michael Gleicher Many motion capture editing operations result in the feet of the character moving when they ought to remain planted. We present a simple, efficient algorithm for removing this footskate.
	Rocket Coaster By: Andrew Selle and Marcin Szymanski Rocket Coaster is a roller coaster simulation written for the Undergraduate Graphics course. It implements some interesting graphics and simulation features. Notably it has stenciled shadows, reflections, particle effects, real-time editing of tracks and terrain, and advanced wheel alignment. It also has people.
	Simplicial Families of Drawings By: Lucas Kovar, Michael Gleicher Starting with a small number of drawings, by answering a series of yes/no questions an artist can build a simplicial complex representing a family of similar drawings. Using this simplicial family, one can generate random drawings, interpolate drawings, project a drawing onto the family, or "tug" on a drawing to modify it.
	HijackGL: Non-invasive Extensibility for Graphics Applications By: Alex Mohr, Michael Gleicher, Christopher Herrman HijackGL is a prototype system that allows us to non-invasively intercept graphics applications and change their behavior on the fly.
	The Prokudin-Gorsky Photographs, Recombined Automatically By: Russell Manning Here is demonstrated an automatic technique for recombining the separated color channels of Prokudin-Gorsky's photographs to recreate the original color images. All 1903 photos in the Library of Congress collection are shown.
	Non-photorealistic Rendering Quake By: Alex Mohr, Erik Bakke, Andrew Gardner, Christopher Herrman, Steve Dutcher NPRQuake attempts to capture the elements of different drawing techniques, and immerse the viewer in worlds drawn entirely in the prescribed style. If you have ever imagined running around inside a painting or a drawing, you are beginning to get the idea.
	LibTarga By: Alex Mohr and Mark Pingel This targa library gives the user the ability to easily read and write various (all) TARGA-vision file formats.
	Automatic footplant detection for motion capture data By: Peter Bodiks Finding moments in time when a character's feet collide firmly with the floor are an important part of editing motion capture data. This project explores a number of ways to detect moments of impact between an appendage and an immobile object.
	Visualization of intense fusion reactions By: Mike Wade Large amounts of seemingly patternless data are common with many burgeoning scientific experiments. This project's intent was to do a number of automatic operations on large sets of data to both categorize and refine their results.
	Visualization of the molecular dynamics of the neuron By: Ian P. Coyle Generation of high quality animations using scale models of molecules to illustrate the complex events in neuron function, such as synaptic transmission, axon pathfinding, and learning.
	Still-image transitions By: Novita Aryanis Image-to-image transitions were experimented with in this project.
	From hand-written notes to the web By: Craig Farrand Automation of converting handwritten notes on arbitrary types of paper to publishing final webpages with said data.