A Framework for the Acquisition of Multimodal Human-Robot Interaction Data Sets with a Whole-System Perspective

Titelaufnahme

Titel
A Framework for the Acquisition of Multimodal Human-Robot Interaction Data Sets with a Whole-System Perspective
Verfasser
Wienke, Johannes ; Klotz, David ; Wrede, Sebastian
Herausgeber
Edlund, Jens ; Heylen, Dirk ; Paggio, Patrizia
Erschienen
2012
Sprache
Englisch
Dokumenttyp
Konferenzband
Schlagwörter
human-robot interaction
URN
urn:nbn:de:0070-pub-24870854

Zugriffsbeschränkung

Links

Dateien

Klassifikation

Abstract

In this work we present a conceptual framework for the creation of multimodal data sets which combine human-robot interaction with system-level data from the robot platform. The framework is based on the assumption that perception, interaction modeling and system integration need to be treated jointly in order to improve human-robot interaction capabilities of current robots. To demonstrate the feasibility of the framework, we describe how it has been realized for the recording of a data set with the humanoid robot NAO.

Statistik