News
About
Projects
Contact
Material
Calendar
Resources

Motion tracking with infra-red imaging

Advisors: Ricardo Gutierrez-Osuna, Steve Ortiz, Marco Zavala

Project Description:

The objective of this project is to develop a robust tracking system for a facial motion-capture apparatus. Based on IBM's pupilCam design, we have developed an infra-red (IR) audio-visual capture that is capable of tracking, in real-time, a number of retro-reflective markers temporarily glued to the face of a speaker. The current system is functional, but requires a major upgrade before it can be reliably used to acquire large amounts of audio-visual data (minutes to hours).

Educational Objectives:

  1. Low-level computer vision and speech processing
  2. Human-computer interaction: multi-modal interfaces
  3. System-integration: signal processing, audio-visual data acquisition, graphical user interfaces

Deliverables:

  1. User-friendly and efficient GUI for audio-visual capture
  2. Implementation of a robust racking algorithm, integrated with the GUI
  3. Project documentation and live demonstration, as described in the course syllabus