A monocular camera-based robot leader-follower system: Camille Monnier, Charles River Analytics Inc.

Starts:
1:00 pm on Monday, April 1, 2013
Ends:
2:00 pm on Monday, April 1, 2013
Location:
MCS 148
Abstract: Current generation control systems for mobile robots and unmanned ground vehicles (UGV) typically require operators to physically control a platform through teleoperation, even for simple tasks such as travelling from one location to another. This is particularly cumbersome if the UGV is supposed to act as a mule and simply follow its leader. While vision-based control technologies promise to significantly reduce the burden on UGV operators, most schemes rely on specialized sensing hardware, such as LIDAR or stereo cameras, or require additional markers to differentiate the leader from nearby pedestrians. We present a system for leader-follower control of small UGVs using only a single monocular camera, which is ubiquitous on mobile platforms. The system allows a user to control a mobile robot by leading the way and issuing commands through arm/hand gestures, and differentiates between the leader and nearby bystanders. Our system achieves this through efficient extraction and re-use of features across the pedestrian detection, appearance learning, and gesture recognition algorithms. In this talk, I will discuss our approach to person detection, tracking, and gesture recognition, and discuss some of our results and considerations for real-world systems. Bio: Camille Monnier is a Senior Scientist at Charles River Analytics, where he leads research and development of vision systems for unmanned systems and security and surveillance. At Charles River, Mr. Monnier is the architect of a generalized object detection and classification software library supporting a variety of applications, including vision-based navigation for ground, surface, and underwater vehicles; human-computer interaction (HCI); and intelligence, surveillance, and reconnaissance (ISR). His research interests include object detection and tracking, human pose estimation, and feature extraction. He holds a Bachelor's degree in Physics from Boston University.