Gesture recognition in the haptic creature

Authors:
Jonathan Chang;Karon MacLean;Steve Yohanan
Affiliations:
Department of Computer Science, University of British Columbia, Vancouver, B.C., Canada;Department of Computer Science, University of British Columbia, Vancouver, B.C., Canada;Department of Computer Science, University of British Columbia, Vancouver, B.C., Canada
Venue:
EuroHaptics'10 Proceedings of the 2010 international conference on Haptics: generating and perceiving tangible sensations, Part I
Year:
2010

Citing 3
Cited 4

A tool to study affective touch

CHI '09 Extended Abstracts on Human Factors in Computing Systems
Artificial Intelligence: A Modern Approach

Artificial Intelligence: A Modern Approach
Real-time social touch gesture recognition for sensate robots

IROS'09 Proceedings of the 2009 IEEE/RSJ international conference on Intelligent robots and systems

Design and assessment of the haptic creature's affect display

Proceedings of the 6th international conference on Human-robot interaction
Detecting shape deformation of soft objects using directional photoreflectivity measurement

Proceedings of the 24th annual ACM symposium on User interface software and technology
Affective touch gesture recognition for a furry zoomorphic machine

Proceedings of the 7th International Conference on Tangible, Embedded and Embodied Interaction
Clustering approach to characterize haptic expressions of emotions

ACM Transactions on Applied Perception (TAP)

Quantified Score

Hi-index	0.00

Visualization

Abstract

Touch is an important but poorly studied aspect of emotional communication. With the Haptic Creature we are investigating fundamentals of affective touch. This small robot senses the world solely by being touched via a force-sensing resistor network, and communicates its internal state via purring, stiffening its ears and modulating its breathing and pulse. We describe the Creature's first-generation gesture recognition engine, analyze its results, and specify its next iteration. In the region of highest sensor density, four gestures were differentiated with an average of 77% accuracy. Error patterns suggest that sensor deficiency rather than algorithm pose current performance limits.