une

Partenaires

CNRS
Logo tutelle
Logo carnot


Rechercher

Sur ce site


Accueil > À noter > Séminaires > Mardi 4 Avril 2017, Guido DE CROON (Delft University of Technology, Pays-Bas). A 17h45, en Salle des thèses (A104), Faculté des Sciences du Sport, Campus de Luminy, Marseille. ’Optical-Flow based Self-Supervised Learning of Obstacle Appearance applied to MAV Landing’.

Mardi 4 Avril 2017, Guido DE CROON (Delft University of Technology, Pays-Bas). A 17h45, en Salle des thèses (A104), Faculté des Sciences du Sport, Campus de Luminy, Marseille. ’Optical-Flow based Self-Supervised Learning of Obstacle Appearance applied to MAV Landing’.

Invité par Franck RUFFIER

Mise à jour : 24 mars

Monocular motion cues have been widely used by Micro Air Vehicles (MAVs) to detect obstacles during visual navigation. However, this approach requires significant movement, which reduces the efficiency of navigation and may even introduce risks in narrow spaces. In this talk, I will present a novel setup of self-supervised learning (SSL), in which motion cues serve as a scaffold to learn the visual appearance of obstacles in the environment. We applied this learning method to a landing task, in which initially ’surface roughness’ is estimated from the optical flow field in order to detect obstacles. Subsequently, a regression function is learned that maps appearance features represented by texton distributions to a surface roughness estimate. After learning, the MAV can detect obstacles by just analyzing a still image. This allows the MAV to search for a landing spot without moving. We first demonstrated this principle to work with offline tests involving images captured from an on-board camera, and then demonstrated the principle in flight. Although surface roughness is a property of the entire flow field in the global image, the appearance learning even allows for the pixel-wise segmentation of obstacles.