de
en
Schliessen
Detailsuche
Bibliotheken
Projekt
Impressum
Datenschutz
zum Inhalt
Detailsuche
Schnellsuche:
OK
Ergebnisliste
Titel
Titel
Inhalt
Inhalt
Seite
Seite
Im Dokument suchen
Lohan, Katrin Solveig: A model of contingency detection to spot tutoring behavior and respond to ostensive cues in human-robot-interaction. 2011
Inhalt
Acknowledgments
Contents
List of Figures
List of Tables
Acronyms
1 Introduction
1.1 Motivation
1.2 Structure of the presented work
2 Insights from developmental psychology
2.1 Related Work: Attention Mechanisms and Features in a Tutoring Situation
2.1.1 Infants' benefit from tutoring behavior
2.1.2 Attention is needed to create a social interaction
2.1.3 Children prefer contingent actions
2.1.4 Summary
2.2 Related Work: Ostensive signals guide attention
2.2.1 Ostensive signals highlighting novel information
2.2.2 Innate structuring mechanism (contingency)
2.2.3 Motionese: a polemic hand motion to guide infant's attention
2.2.4 Child directed speech scaffolds infants (motherese)
2.2.5 Children prefer faces
2.2.6 Summary
2.3 Analysis: Crucial factors for a successful interaction
2.3.1 Motionese Corpus
2.3.2 Looming: a way of making a movement salient
2.3.3 Learners' age affects the tutoring behavior
2.3.4 Infants anticipate the tutor's actions
2.3.5 Summary
3 Insights from Human Robot Interaction
3.1 Related Work: Using saliency to control robotic gazing behavior
3.1.1 Creating attention with a saliency system
3.1.2 The iCub humanoid robot
3.1.3 Summary
3.2 Analysis: Attention in a Tutoring Situation with a Robot
3.2.1 Robot-Directed Interaction Experiment (RDIE)
3.2.2 Analysing the different dimensions of behavior adaptation
3.2.3 Do the measures for ostensive signals generalize over different tasks?
3.2.4 Embodiment Corpus
3.2.5 Analysis if embodiment's effect on tutoring behavior
3.2.6 Can state-of-the-art saliency systems model infant gazing behavior in tutoring situations?
3.2.7 Summary
4 A new model for detecting a tutoring situation
4.1 Related Work: Contingency in other work
4.1.1 An Infomax controller for real time detection of social contingency
4.1.2 Vision-based contingency detection
4.1.3 Summary
4.2 Modeling: Deriving a feature set for a tutoring spotter
4.2.1 Model of contingency detection
4.2.2 Modeling Behavior
4.2.3 Implementation of the model
4.2.4 Summary
4.3 Analysis: The tutoring spotter in a interaction with a human tutor
4.3.1 Studying the tutoring spotter
4.3.2 Random behavior
4.3.3 Analysing the implemented contingency behavior
4.3.4 Questionnaire
4.3.5 Summary
4.4 Discussion and further directions
5 Revising the Tutor Spotter
5.1 Contingency Corpus
5.1.1 Gaze direction detection
5.1.2 Summary
5.2 Evaluation: The tutoring spotter 2.0
5.2.1 Eye gaze Results
5.2.2 Contingency Results
5.2.3 Speech results for the cups stacking task
5.2.4 Questionnaires Results
5.2.5 Discussion
5.3 Analysis: Further promising feedback strategies
5.3.1 Anticipation as a feedback mechanism
5.3.2 Keyword spotting
5.3.3 Levels of embodiment: linguistic analysis of factors influencing HRI
5.3.4 Summary
6 Conclusion
6.1 Summary
6.2 Capability and Problems
6.3 Future Work
Appendix
A Appendix
A.1 Motionese Corpus
A.1.1 Speech analysis of Manner and Path
A.1.2 Looming
A.2 Contingency Corpus
A.2.1 Questionnaire
A.3 Technical details
A.3.1 Hardware
A.3.2 Software
A.3.3 Statistical tests
Bibliography