Training Lightweight CNNs for Human-Nanodrone Proximity Interaction from Small Datasets using Background Randomization

no code implementations • 27 Oct 2021 • Marco Ferri, Dario Mantegazza, Elia Cereda, Nicky Zimmerman, Luca M. Gambardella, Daniele Palossi, Jérôme Guzzi, Alessandro Giusti

We consider the task of visually estimating the pose of a human from images acquired by a nearby nano-drone; in this context, we propose a data augmentation approach based on synthetic background substitution to learn a lightweight CNN model from a small real-world training set.

Data Augmentation