End-to-end Reinforcement Learning for Time Optimal Quadcopter Flight @microuav
End-to-end Reinforcement Learning for Time Optimal Quadcopter Flight  @microuav
Uploaded March 2024 | Updated November 2026, 3 hours ago
We introduce end-to-end reinforcement learning for optimal quadcopter flight. Our network directly outputs motor commands and is compared to a traditional method where the network interfaces with an INDI controller. In simulation our method excels due to its direct control authority. In real-world tests, although the performance gap narrows, our method maintains a slight advantage, demonstrating the potential of end-to-end algorithms.
End-to-end Reinforcement Learning for Time Optimal Quadcopter FlightIMAV 2014 Presentations - Day 2Morpheus MK1 UA onboardcamSelf-supervised monocular distance learning on an 18-gram DelFlySwarm exploration by tiny flying robots - short versionARdrone RC controledNanoFlowNet: Real-Time Dense Optical Flow on a Nano QuadcopterIMAV 2014 Presentations - Fabrizio SchianoIMAV2015 MAVlab TUDelftTU Delft Quadshot - Proof of FlightEnd-to-end Neural Network Based Optimal Quadcopter ControlIMAV 2022 - Greenhouse Challenge
MAVLab TU Delft |

End-to-end Reinforcement Learning for Time Optimal Quadcopter Flight

SHARE TO X SHARE TO REDDIT SHARE TO FACEBOOK WALLPAPER