A neural network based approach to social touch classification

Siewart van Wingerden, Tobias J. Uebbing, Merel Madeleine Jung, Mannes Poel

Research output: Chapter in Book/Report/Conference proceedingConference contribution

  • 5 Citations

Abstract

Touch is an important interaction modality in social interaction, for instance touch can communicate emotions and can intensify emotions communicated by other modalities. In this paper we explore the use of Neural Networks for the classification of touch. The exploration and assessment of Neural Networks (NNs) is based on the Corpus of Social Touch established by Jung et al. This corpus was split in a train set (65%) and test set (35%), the train set was used to find the optimal parameters for the NN and for training the final model. Also different feature sets were investigated; the basic feature set included in the corpus, energy-histogram and dynamical features. Using all features led to the best performance of 64% on the test set, using a NN consisting of one hidden layer with 46 neurones. The confusion matrix showed the expected high confusion between pat-tap and grab-squeeze. A leave-one-subject-out approach lead to a performance of 54%, which is comparable with the results of Jung et al.
LanguageUndefined
Title of host publicationProceedings of the 2nd International Workshop on Emotion representations and modelling in Human-Computer Interaction systems, ERM4HCI 2014
Place of PublicationNew York
PublisherAssociation for Computing Machinery
Pages7-12
Number of pages6
ISBN (Print)978-1-4503-0124-4
DOIs
StatePublished - Nov 2014

Publication series

NameERM4HCI '14
PublisherACM

Keywords

  • EWI-25281
  • Social Touch
  • METIS-309649
  • IR-93289
  • Neural Networks
  • Touch gesture recognition

Cite this

van Wingerden, S., Uebbing, T. J., Jung, M. M., & Poel, M. (2014). A neural network based approach to social touch classification. In Proceedings of the 2nd International Workshop on Emotion representations and modelling in Human-Computer Interaction systems, ERM4HCI 2014 (pp. 7-12). (ERM4HCI '14). New York: Association for Computing Machinery. DOI: 10.1145/2668056.2668060
van Wingerden, Siewart ; Uebbing, Tobias J. ; Jung, Merel Madeleine ; Poel, Mannes. / A neural network based approach to social touch classification. Proceedings of the 2nd International Workshop on Emotion representations and modelling in Human-Computer Interaction systems, ERM4HCI 2014. New York : Association for Computing Machinery, 2014. pp. 7-12 (ERM4HCI '14).
@inproceedings{3c658b6637254357a7b54b53accd0234,
title = "A neural network based approach to social touch classification",
abstract = "Touch is an important interaction modality in social interaction, for instance touch can communicate emotions and can intensify emotions communicated by other modalities. In this paper we explore the use of Neural Networks for the classification of touch. The exploration and assessment of Neural Networks (NNs) is based on the Corpus of Social Touch established by Jung et al. This corpus was split in a train set (65{\%}) and test set (35{\%}), the train set was used to find the optimal parameters for the NN and for training the final model. Also different feature sets were investigated; the basic feature set included in the corpus, energy-histogram and dynamical features. Using all features led to the best performance of 64{\%} on the test set, using a NN consisting of one hidden layer with 46 neurones. The confusion matrix showed the expected high confusion between pat-tap and grab-squeeze. A leave-one-subject-out approach lead to a performance of 54{\%}, which is comparable with the results of Jung et al.",
keywords = "EWI-25281, Social Touch, METIS-309649, IR-93289, Neural Networks, Touch gesture recognition",
author = "{van Wingerden}, Siewart and Uebbing, {Tobias J.} and Jung, {Merel Madeleine} and Mannes Poel",
note = "10.1145/2668056.2668060",
year = "2014",
month = "11",
doi = "10.1145/2668056.2668060",
language = "Undefined",
isbn = "978-1-4503-0124-4",
series = "ERM4HCI '14",
publisher = "Association for Computing Machinery",
pages = "7--12",
booktitle = "Proceedings of the 2nd International Workshop on Emotion representations and modelling in Human-Computer Interaction systems, ERM4HCI 2014",
address = "United States",

}

van Wingerden, S, Uebbing, TJ, Jung, MM & Poel, M 2014, A neural network based approach to social touch classification. in Proceedings of the 2nd International Workshop on Emotion representations and modelling in Human-Computer Interaction systems, ERM4HCI 2014. ERM4HCI '14, Association for Computing Machinery, New York, pp. 7-12. DOI: 10.1145/2668056.2668060

A neural network based approach to social touch classification. / van Wingerden, Siewart; Uebbing, Tobias J.; Jung, Merel Madeleine; Poel, Mannes.

Proceedings of the 2nd International Workshop on Emotion representations and modelling in Human-Computer Interaction systems, ERM4HCI 2014. New York : Association for Computing Machinery, 2014. p. 7-12 (ERM4HCI '14).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - A neural network based approach to social touch classification

AU - van Wingerden,Siewart

AU - Uebbing,Tobias J.

AU - Jung,Merel Madeleine

AU - Poel,Mannes

N1 - 10.1145/2668056.2668060

PY - 2014/11

Y1 - 2014/11

N2 - Touch is an important interaction modality in social interaction, for instance touch can communicate emotions and can intensify emotions communicated by other modalities. In this paper we explore the use of Neural Networks for the classification of touch. The exploration and assessment of Neural Networks (NNs) is based on the Corpus of Social Touch established by Jung et al. This corpus was split in a train set (65%) and test set (35%), the train set was used to find the optimal parameters for the NN and for training the final model. Also different feature sets were investigated; the basic feature set included in the corpus, energy-histogram and dynamical features. Using all features led to the best performance of 64% on the test set, using a NN consisting of one hidden layer with 46 neurones. The confusion matrix showed the expected high confusion between pat-tap and grab-squeeze. A leave-one-subject-out approach lead to a performance of 54%, which is comparable with the results of Jung et al.

AB - Touch is an important interaction modality in social interaction, for instance touch can communicate emotions and can intensify emotions communicated by other modalities. In this paper we explore the use of Neural Networks for the classification of touch. The exploration and assessment of Neural Networks (NNs) is based on the Corpus of Social Touch established by Jung et al. This corpus was split in a train set (65%) and test set (35%), the train set was used to find the optimal parameters for the NN and for training the final model. Also different feature sets were investigated; the basic feature set included in the corpus, energy-histogram and dynamical features. Using all features led to the best performance of 64% on the test set, using a NN consisting of one hidden layer with 46 neurones. The confusion matrix showed the expected high confusion between pat-tap and grab-squeeze. A leave-one-subject-out approach lead to a performance of 54%, which is comparable with the results of Jung et al.

KW - EWI-25281

KW - Social Touch

KW - METIS-309649

KW - IR-93289

KW - Neural Networks

KW - Touch gesture recognition

U2 - 10.1145/2668056.2668060

DO - 10.1145/2668056.2668060

M3 - Conference contribution

SN - 978-1-4503-0124-4

T3 - ERM4HCI '14

SP - 7

EP - 12

BT - Proceedings of the 2nd International Workshop on Emotion representations and modelling in Human-Computer Interaction systems, ERM4HCI 2014

PB - Association for Computing Machinery

CY - New York

ER -

van Wingerden S, Uebbing TJ, Jung MM, Poel M. A neural network based approach to social touch classification. In Proceedings of the 2nd International Workshop on Emotion representations and modelling in Human-Computer Interaction systems, ERM4HCI 2014. New York: Association for Computing Machinery. 2014. p. 7-12. (ERM4HCI '14). Available from, DOI: 10.1145/2668056.2668060