Collision-Free UAV Navigation with a Monocular Camera Using Deep Reinforcement Learning.
Yun ChenNuria González PrelcicRobert W. Heath Jr.Published in: MLSP (2020)
Keyphrases
- collision free
- monocular camera
- path planning
- reinforcement learning
- visual odometry
- autonomous navigation
- dynamic environments
- mobile robot
- motion planning
- collision avoidance
- feature tracking
- inertial sensors
- ground plane
- real time
- multi robot
- multibody
- multiple moving objects
- indoor environments
- path finding
- optimal policy
- free space
- optimal path
- state space
- control policy
- degrees of freedom
- moving objects
- optimal control
- simultaneous localization and mapping
- multi modal
- motion tracking
- feature detection
- vision system
- optical flow
- structure from motion
- computer vision