Fast Traversability Estimation for Wild Visual Navigation
Jonas Frey, Matías Mattamala, Nived Chebrolu, César Cadena, Maurice Fallon, Marco Hutter
Abstract
Fig. 1: Wild Visual Navigation (WVN) learns to predict traversability from images via online self-supervised learning.Starting from a randomly initialized traversability estimation network without prior assumptions about the environment (a), a human operator drives the robot around areas that are traversable for the given platform (b).After a few minutes of operation, WVN learns to distinguish between traversable and untraversable areas (c), enabling the robot to navigate autonomously and safely within the environment (d).
Topics & Concepts
Computer visionEstimationComputer scienceArtificial intelligenceEngineeringSystems engineeringVideo Surveillance and Tracking MethodsRobotic Path Planning AlgorithmsRobotics and Sensor-Based Localization