Abstract :  Recent advances in Deep Learning led to remarkable solutions in a wide range of applications. Deep Learning-empowered systems can nowadays achieve performance levels in various Computer Vision tasks which are comparable to, or even exceeding, that of humans. Even though these advancements have the potential to open new high-impact applications in Robotics, this promise has yet to be met. This is due to challenges in Robotics which go beyond the unrestricted analysis of images. Autonomous robots need to perform multiple analyses for understanding their environments and navigate inside it, under restrictions concerning real-time operation and computational/energy power. Addressing these challenges is crucial for various types  of robotic platforms, including low power Unmanned Aerial Vehicles (drones), Collaborative Robots, and Autonomous Vehicles. The purpose of the Special Session is to provide a forum to exchange ideas and to discuss developments in Deep Learning models for Robotic Perception and Navigation. The Special Session is supported by the H2020 project OpenDR.

Alexandros Iosifidis
Aarhus University
Denmark

Anastasios Tefas
Aristotle University of Thessaloniki
Greece