CSAIL Event Calendar: Previous Series
Reducing Drift in Parametric Motion Tracking
Speaker: Ali Rahimi , Vision Interface Project - MIT AI Laboratory
In this talk, I will describe a simple trick for reducing drift in vision- based trackers. Modern techniques for tracking the location of objects range from GPS and magnetic field sensing, to sonar and vision-based techniques. Traditionally, most of these techniques are either inaccurate, limited to a few domains, or suffer from drift. Accurately tracking an object for prolonged periods of time is difficult no matter what technology is used. Most trackers use a Markov chain as their measurement model, and use a Kalman filter to compute optimal pose estimates. I will review some tracking techniques and explain the limitations of this model. I will then present a strategy for improving the performance of vision-based trackers, and show how to take full advantage of this idea by using a graphical model to solve for the optimal pose of an object in a video sequence.