Robots with Commonsense: Improving Object Recognition through Size and Spatial Awareness

Chiatti, Agnese; Motta, Enrico and Daga, Enrico Robots with Commonsense: Improving Object Recognition through Size and Spatial Awareness. In: the AAAI 2022 Spring Symposium on Machine Learning and Knowledge Engineering for Hybrid Intelligence (AAAI-MAKE 2022) (Martin, Andreas; Hinkelmann, Knut; Fill, Hans-Georg; Gerber, Aurona; Lenat, Doug; Stolle, Reinhard and van Harmelen, Frank eds.), 21-23 Mar 2022, Palo Alto, California, USA, CEUR.

URL: https://www.aaai-make.info/

Abstract

To effectively assist us with our daily tasks, service robots need object recognition methods that perform robustly in dynamic environments. Our prior work has shown that augmenting Deep Learning (DL) methods with knowledge-based reasoning can drastically improve the reliability of object recognition systems. This paper proposes a novel method to equip DL-based object recognition with the ability to reason on the typical size and spatial relations of objects. Experiments in a real-world robotic scenario show that the proposed hybrid architecture significantly outperforms DL-only solutions.

Viewing alternatives

Download history

Item Actions

Export

About

Recommendations