Deep Learning for Emotion Recognition in Faces

Ariel Ruiz-Garcia, Mark Elshaw, Abdulrahman Altahhan, Vasile Palade

Research output: Chapter in Book/Report/Conference proceedingChapter

4 Citations (Scopus)

Abstract

Deep Learning (DL) has shown real promise for the classification efficiency for emotion recognition problems. In this paper we present experimental results for a deeply-trained model for emotion recognition through the use of facial expression images. We explore two Convolutional Neural Network (CNN) architectures that offer automatic feature extraction and representation, followed by fully connected softmax layers to classify images into seven emotions. The first architecture explores the impact of reducing the number of deep learning layers and the second splits the input images horizontally into two streams based on eye and mouth positions. The first proposed architecture produces state of the art results with an accuracy rate of 96.93 % and the second architecture with split input produces an average accuracy rate of 86.73 %, respectively.
Original languageEnglish
Title of host publicationArtificial Neural Networks and Machine Learning – ICANN 2016
EditorsAlessandro E.P. Villa, Paolo Masulli, Antonio Javier Pons Rivero
Place of PublicationSwitzerland
PublisherSpringer Verlag
Pages38-46
Volume9887
ISBN (Print)978-3-319-44780-3, 978-3-319-44781-0
DOIs
Publication statusPublished - 13 Aug 2016
EventThe 25th International Conference on Artificial Neural Networks - Barcelona, Spain
Duration: 6 Sep 20169 Sep 2016

Conference

ConferenceThe 25th International Conference on Artificial Neural Networks
Abbreviated titleICANN 2016
CountrySpain
CityBarcelona
Period6/09/169/09/16

Fingerprint

Network architecture
Feature extraction
Neural networks
Deep learning

Bibliographical note

The full text is not available on the repository.

Keywords

  • Deep learning
  • Convolution neural networks
  • Emotion recognition
  • Empathic robots

Cite this

Ruiz-Garcia, A., Elshaw, M., Altahhan, A., & Palade, V. (2016). Deep Learning for Emotion Recognition in Faces. In A. E. P. Villa, P. Masulli, & A. J. P. Rivero (Eds.), Artificial Neural Networks and Machine Learning – ICANN 2016 (Vol. 9887, pp. 38-46). Switzerland: Springer Verlag. https://doi.org/10.1007/978-3-319-44781-0_5

Deep Learning for Emotion Recognition in Faces. / Ruiz-Garcia, Ariel; Elshaw, Mark; Altahhan, Abdulrahman; Palade, Vasile.

Artificial Neural Networks and Machine Learning – ICANN 2016. ed. / Alessandro E.P. Villa; Paolo Masulli; Antonio Javier Pons Rivero. Vol. 9887 Switzerland : Springer Verlag, 2016. p. 38-46.

Research output: Chapter in Book/Report/Conference proceedingChapter

Ruiz-Garcia, A, Elshaw, M, Altahhan, A & Palade, V 2016, Deep Learning for Emotion Recognition in Faces. in AEP Villa, P Masulli & AJP Rivero (eds), Artificial Neural Networks and Machine Learning – ICANN 2016. vol. 9887, Springer Verlag, Switzerland, pp. 38-46, The 25th International Conference on Artificial Neural Networks, Barcelona, Spain, 6/09/16. https://doi.org/10.1007/978-3-319-44781-0_5
Ruiz-Garcia A, Elshaw M, Altahhan A, Palade V. Deep Learning for Emotion Recognition in Faces. In Villa AEP, Masulli P, Rivero AJP, editors, Artificial Neural Networks and Machine Learning – ICANN 2016. Vol. 9887. Switzerland: Springer Verlag. 2016. p. 38-46 https://doi.org/10.1007/978-3-319-44781-0_5
Ruiz-Garcia, Ariel ; Elshaw, Mark ; Altahhan, Abdulrahman ; Palade, Vasile. / Deep Learning for Emotion Recognition in Faces. Artificial Neural Networks and Machine Learning – ICANN 2016. editor / Alessandro E.P. Villa ; Paolo Masulli ; Antonio Javier Pons Rivero. Vol. 9887 Switzerland : Springer Verlag, 2016. pp. 38-46
@inbook{fbf8e924c9b448a6a129415d58b78dd5,
title = "Deep Learning for Emotion Recognition in Faces",
abstract = "Deep Learning (DL) has shown real promise for the classification efficiency for emotion recognition problems. In this paper we present experimental results for a deeply-trained model for emotion recognition through the use of facial expression images. We explore two Convolutional Neural Network (CNN) architectures that offer automatic feature extraction and representation, followed by fully connected softmax layers to classify images into seven emotions. The first architecture explores the impact of reducing the number of deep learning layers and the second splits the input images horizontally into two streams based on eye and mouth positions. The first proposed architecture produces state of the art results with an accuracy rate of 96.93 {\%} and the second architecture with split input produces an average accuracy rate of 86.73 {\%}, respectively.",
keywords = "Deep learning, Convolution neural networks, Emotion recognition, Empathic robots",
author = "Ariel Ruiz-Garcia and Mark Elshaw and Abdulrahman Altahhan and Vasile Palade",
note = "The full text is not available on the repository.",
year = "2016",
month = "8",
day = "13",
doi = "10.1007/978-3-319-44781-0_5",
language = "English",
isbn = "978-3-319-44780-3",
volume = "9887",
pages = "38--46",
editor = "Villa, {Alessandro E.P.} and Paolo Masulli and Rivero, {Antonio Javier Pons}",
booktitle = "Artificial Neural Networks and Machine Learning – ICANN 2016",
publisher = "Springer Verlag",
address = "Austria",

}

TY - CHAP

T1 - Deep Learning for Emotion Recognition in Faces

AU - Ruiz-Garcia, Ariel

AU - Elshaw, Mark

AU - Altahhan, Abdulrahman

AU - Palade, Vasile

N1 - The full text is not available on the repository.

PY - 2016/8/13

Y1 - 2016/8/13

N2 - Deep Learning (DL) has shown real promise for the classification efficiency for emotion recognition problems. In this paper we present experimental results for a deeply-trained model for emotion recognition through the use of facial expression images. We explore two Convolutional Neural Network (CNN) architectures that offer automatic feature extraction and representation, followed by fully connected softmax layers to classify images into seven emotions. The first architecture explores the impact of reducing the number of deep learning layers and the second splits the input images horizontally into two streams based on eye and mouth positions. The first proposed architecture produces state of the art results with an accuracy rate of 96.93 % and the second architecture with split input produces an average accuracy rate of 86.73 %, respectively.

AB - Deep Learning (DL) has shown real promise for the classification efficiency for emotion recognition problems. In this paper we present experimental results for a deeply-trained model for emotion recognition through the use of facial expression images. We explore two Convolutional Neural Network (CNN) architectures that offer automatic feature extraction and representation, followed by fully connected softmax layers to classify images into seven emotions. The first architecture explores the impact of reducing the number of deep learning layers and the second splits the input images horizontally into two streams based on eye and mouth positions. The first proposed architecture produces state of the art results with an accuracy rate of 96.93 % and the second architecture with split input produces an average accuracy rate of 86.73 %, respectively.

KW - Deep learning

KW - Convolution neural networks

KW - Emotion recognition

KW - Empathic robots

U2 - 10.1007/978-3-319-44781-0_5

DO - 10.1007/978-3-319-44781-0_5

M3 - Chapter

SN - 978-3-319-44780-3

SN - 978-3-319-44781-0

VL - 9887

SP - 38

EP - 46

BT - Artificial Neural Networks and Machine Learning – ICANN 2016

A2 - Villa, Alessandro E.P.

A2 - Masulli, Paolo

A2 - Rivero, Antonio Javier Pons

PB - Springer Verlag

CY - Switzerland

ER -