Danette Allen's Ph.D. Research Progress
The Wide Field of View HiBall, "WFOV-HiBall", is the second generation of the HiBall
designed and developed at UNC-CH. We hope to develop a similar system
that reduces the infrastructure requirements (fewer LEDs,
etc.) and shrinks the sensor unit by 1/3 to 1/2. In order to accomplish
this, we need an analysis tool. From this, my Ph.D. dissertation topic
was born. Artemis is a system for the analysis and comparison of the
estimated performance of tracking and motion capture systems. It consists of two parts: a mathematical
framework and an interactive visualization environment. The goal of Artemis is to provide
designers with a tool that affords insight into the performance effects of
candidate tracker design choices in the context of the desired working volume,
the expected user motion dynamics, and other concerns.
Thesis Statement (June 2001)
A framework employing stochastic state-space analysis and appropriate
visualization techniques offers new insight into the design of tracking systems
and improves the design process. Such a framework allows interaction with the
design parameters, enabling comprehension of overall performance tradeoffs
prior to final product characterization without requiring path-dependent
simulation.
Proposal (available only to cs.unc.edu)
- First Draft January 21, 2001.
- Second Draft February 09, 2001.
- Draft 2.1 February 12, 2001.
- Draft 2.2 February 13, 2001.
- Draft 2.3 March 06, 2001.
- Third Draft March 26, 2001.
- Draft 3.1 April 03, 2001.
- Fourth Draft April 12, 2001.
- Draft 4.1 April 18, 2001.
- Draft 4.2 April 26, 2001.
- Final Version June 11, 2001
My Integrative Paper (IP) , "A Tracker Performance Analysis Tool" (final draft November 24, 2001). This is available only to cs.unc.edu.
I am a member of the UNC Tracker Research Group. I helped develop and present a course at SIGGRAPH 2001: Tracking: Beyond 15 Minutes of Thought (Course 11) with Gary Bishop and Greg Welch.