[Apr. 2nd] Seminar by Mr. Dan Stronger (The University of Texas at Austin)
Autonomous Sensor and Action Model Learning for Mobile Robots
- Speaker: Mr. Dan Stronger
- Location: CH 430
- Date: Wednesday, April 2
- Time: 11:00 am - 12:00 pm
Seminar Abstract
In order for a mobile robot to accurately interpret its sensations and
predict the effects of its decisions, it must have accurate models of its
sensors and actions. These models are typically tuned manually, a brittle
and laborious process. Autonomous model learning is a promising
alternative to manual calibration, but previous work has assumed the
presence of an accurate action or sensor model in order to train the other
model. This talk presents a novel methodology to enable mobile robots to
learn both their action and sensor models, starting without an accurate
version of either. This methodology is based on an adaptation of the
Expectation-Maximization (EM) algorithm, where the learned parameters are
the action and sensor models. The resulting technique is validated
experimentally on a Sony Aibo ERS-7 robot.
Speaker Biography
Daniel Stronger received his Bachelors degree from Harvard University in
2001 and is a Computer Science Ph.D. Candidate at the University of Texas
at Austin. His research is in the area of Artificial Intelligence and
Robotics. Specifically, his primary research focus is the development of
algorithms that enable autonomous mobile robots to learn models of their
actions and sensors with as little human supervision as possible.
Attachments
|
