2007 · 31 citations · 8 references
EngineeringAutonomous SystemsComputer ScienceAutonomous VehiclesRobot LearningRoboticsAutonomous NavigationImitation LearningLearning From DemonstrationNavigation LearningRobot NavigationTeleoperationHuman Operator ModelingImitation Learning For NavigationLearning From TeleoperationModel Learning From DemonstrationsRobot Learning From DemonstrationUnmanned Ground Vehicles
Driving in unknown natural outdoor terrain is a challenge for autonomous ground vehicles. It can be difficult for a robot to discern obstacles and other hazards in its environment, and characteristics of this high cost terrain may change from one environment to another, or even with different lighting conditions. One successful approach to this problem is for a robot to learn from a demonstration by a human operator. In this paper, we describe an approach to calculating terrain costs from Bayesian estimates using feature vectors measured during a short teleoperated training run in similar terrain and conditions. We describe the theory, its implementation on two different robotic systems, and results of several independently conducted field tests.
8
Nathan Ratliff, J. Andrew Bagnell, Martin Zinkevich · 2006 · 640 citations
Off-Road Obstacle Avoidance through End-to-End Learning
Urs Müller, Jan Ben, Eric Cosatto et al. · Neural Information Processing Systems · 2005 · 446 citations
The DARPA PerceptOR evaluation experiments
Eric Krotkov, Scott Fish, Larry Jackel et al. · Autonomous Robots · 2006 · 47 citations