Dario Mantegazza, Jérôme Guzzi, Luca M. Gambardella and Alessandro Giusti
Dalle Molle Institute for Artificial Intelligence (IDSIA), USI-SUPSI, Lugano, Switzerland
We consider a quadrotor equipped with a forward-facing camera, and an user freely moving in its proximity; we control the quadrotor in order to stay in front of the user, using only camera frames. To do so, we train a deep neural network to predict the drone controls given the camera image. Training data is acquired by running a simple hand-designed controller which relies on optical motion tracking data.
The video accepted at HRI2019 is available here.
Other videos are available here
A detailed description of the approach can be found here and on GitHub