The affordance theory provides a biology-inspired approach to enable a robot to act, think and develop like human beings. Based on existing affordance relationships, a robot understands its environment and task in terms of potential actions that it can execute. Deep learning makes it possible for a robot to perceive the environment in an efficient manner. As a result, affordance-based perception together with deep learning provides a possible solution for a robot to provide good service to us. However, affordance knowledge can not be gained just by visual perception and a single object might have multiply affordances. In this paper, we propose a novel framework to combine affordance knowledge and visual perception. Our method has the following features: (i) map human instructions into affordance knowledge; (ii) perceive the environment based on deep neural networks and associate each object with its affordances. In our experiments, a humanoid robot NAO is used and the results demonstrate that affordance knowledge can improve robotic understanding based on deep learning.
Using Affordances to Improve Robotic Understanding Based on Deep Learning
Lect. Notes Electrical Eng.
International Conference on Autonomous Unmanned Systems ; 2021 ; Changsha, China September 24, 2021 - September 26, 2021
Proceedings of 2021 International Conference on Autonomous Unmanned Systems (ICAUS 2021) ; Kapitel : 243 ; 2467-2476
18.03.2022
10 pages
Aufsatz/Kapitel (Buch)
Elektronische Ressource
Englisch
Using Affordances to Improve Robotic Understanding Based on Deep Learning
British Library Conference Proceedings | 2022
|Using Object Affordances to Improve Object Recognition
Deutsches Zentrum für Luft- und Raumfahrt (DLR) | 2011
|Perceiving, learning, and exploiting object affordances for autonomous pile manipulation
British Library Online Contents | 2014
|