A Navigation Engine for Autonomous UAVs with End-to-End Visual Navigation
A 64mW DNN-based Visual Navigation Engine for Autonomous Nano-Drones
We present the first (to the best of our knowledge) demonstration of a navigation engine for autonomous nano-drones capable of closed-loop end-to-end deep convolutional neural network (dnn) based visual navigation.
Our navigation engine is flexible and can be used to span a wide performance range : at its peak performance corner it achieves 18 fps while still consuming on average just 3.5% of the power envelope of the deployed nano-aircraft.
All processing is done with only 64 mw on average.
Daniele Palossi, Antonio Loquercio, Francesco Conti, Eric Flamand, Davide Scaramuzza, Luca Benini