Learning efficient policies for vision-based navigation

Cameras are popular sensors for robot navigation tasks such as localization as they are inexpensive, lightweight, and provide rich data. However, fast movements of a mobile robot typically reduce the performance of vision-based localization systems due to motion blur. In this paper, we present a reinforcement learning approach to choose appropriate velocity profiles for vision-based navigation. The learned policy minimizes the time to reach the destination and implicitly takes the impact of motion blur on observations into account. To reduce the size of the resulting policies, which is desirable in the context of memory-constrained systems, we compress the learned policy via a clustering approach. Extensive simulated and real-world experiments demonstrate that our learned policy significantly outperforms any policy that uses a constant velocity. We furthermore show, that our policy is applicable to different environments. Additional experiments demonstrate that our compressed policies do not result in a performance loss compared to the originally learned policy.

Authors

• Armin Hornung
• Hauke Strasdat
• Maren Bennewitz
• Wolfram Burgard

Keywords

• Navigation
• Robot vision systems
• Cameras
• Robot sensing systems
• Mobile robots
• Learning
• Intelligent robots
• Performance loss
• Unmanned aerial vehicles
• Degradation
• Vision-based Navigation
• Constant Velocity
• Real-world Experiments
• Mobile Robot
• Policy Learning
• Navigation Task
• Motion Blur
• Robot Navigation
• State Space
• Radial Basis Function
• State Representation
• Means Clustering
• Pose Estimation
• Markov Decision Process
• Obstacle Avoidance
• Landmark Localization
• Odometry
• Humanoid Robot
• Simultaneous Localization And Mapping
• Unscented Kalman Filter
• Radial Basis Function Network
• Reinforcement Learning Task
• Real Robot
• Laser Ranging
• Robot Pose
• Target Velocity
• Blur Effect
• Kalman Filter
• Classification Problem