Real-time pain detection in facial expressions for health robotics

Laduona Dai*, Joost Broekens, Khiet Phuong Truong

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

Abstract

Automatic pain detection is an important challenge in health computing. In this paper we report on our efforts to develop a real-time, real-world pain detection system from human facial expressions. Although many studies addressed this challenge, most of them use the same dataset for training and testing. There is no cross-check with other datasets or implementation in real-time to check performance on new data. This is problematic, as evidenced in this paper, because the classifiers overtrain on dataset-specific features. This limits realtime, real-world usage. In this paper, we investigate different methods of real-time pain detection. The training data uses a combination of pain and emotion datasets, unlike other papers. The best model shows an accuracy of 88.4% on a dataset including pain and 7 non-pain emotional expressions. Results suggest that convolutional neural networks (CNN) are not the best methods in some cases as they easily overtrain if the dataset is biased. Finally we implemented our pain detection method on a humanoid robot for physiotherapy. Our work highlights the importance of cross-corpus evaluation & real-time testing, as well as the need for a well balanced and ecologically valid pain dataset.
Original languageEnglish
Title of host publication2019 8th International Conference on Affective Computing and Intelligent Interaction Workshops and Demos (ACIIW)
Place of PublicationPiscataway, NJ
PublisherIEEE
Pages277-283
ISBN (Electronic)978-1-7281-3891-6
ISBN (Print)978-1-7281-3892-3
DOIs
Publication statusPublished - 2019
Event8th International Conference on Affective Computing and Intelligent Interaction Workshops and Demos, ACIIW 2019
- Cambridge, United Kingdom
Duration: 3 Sep 20196 Sep 2019
Conference number: 8

Conference

Conference8th International Conference on Affective Computing and Intelligent Interaction Workshops and Demos, ACIIW 2019
Abbreviated titleACIIW
CountryUnited Kingdom
CityCambridge
Period3/09/196/09/19

Fingerprint

Robotics
Health
Physical therapy
Testing
Classifiers
Robots
Neural networks

Keywords

  • Pain detection
  • Classification
  • Generalization
  • Cross validation
  • Health

Cite this

Dai, L., Broekens, J., & Truong, K. P. (2019). Real-time pain detection in facial expressions for health robotics. In 2019 8th International Conference on Affective Computing and Intelligent Interaction Workshops and Demos (ACIIW) (pp. 277-283). Piscataway, NJ: IEEE. https://doi.org/10.1109/ACIIW.2019.8925192
Dai, Laduona ; Broekens, Joost ; Truong, Khiet Phuong. / Real-time pain detection in facial expressions for health robotics. 2019 8th International Conference on Affective Computing and Intelligent Interaction Workshops and Demos (ACIIW). Piscataway, NJ : IEEE, 2019. pp. 277-283
@inproceedings{b75927e3a70e43ebb06db4f2ffb4c340,
title = "Real-time pain detection in facial expressions for health robotics",
abstract = "Automatic pain detection is an important challenge in health computing. In this paper we report on our efforts to develop a real-time, real-world pain detection system from human facial expressions. Although many studies addressed this challenge, most of them use the same dataset for training and testing. There is no cross-check with other datasets or implementation in real-time to check performance on new data. This is problematic, as evidenced in this paper, because the classifiers overtrain on dataset-specific features. This limits realtime, real-world usage. In this paper, we investigate different methods of real-time pain detection. The training data uses a combination of pain and emotion datasets, unlike other papers. The best model shows an accuracy of 88.4{\%} on a dataset including pain and 7 non-pain emotional expressions. Results suggest that convolutional neural networks (CNN) are not the best methods in some cases as they easily overtrain if the dataset is biased. Finally we implemented our pain detection method on a humanoid robot for physiotherapy. Our work highlights the importance of cross-corpus evaluation & real-time testing, as well as the need for a well balanced and ecologically valid pain dataset.",
keywords = "Pain detection, Classification, Generalization, Cross validation, Health",
author = "Laduona Dai and Joost Broekens and Truong, {Khiet Phuong}",
year = "2019",
doi = "10.1109/ACIIW.2019.8925192",
language = "English",
isbn = "978-1-7281-3892-3",
pages = "277--283",
booktitle = "2019 8th International Conference on Affective Computing and Intelligent Interaction Workshops and Demos (ACIIW)",
publisher = "IEEE",
address = "United States",

}

Dai, L, Broekens, J & Truong, KP 2019, Real-time pain detection in facial expressions for health robotics. in 2019 8th International Conference on Affective Computing and Intelligent Interaction Workshops and Demos (ACIIW). IEEE, Piscataway, NJ, pp. 277-283, 8th International Conference on Affective Computing and Intelligent Interaction Workshops and Demos, ACIIW 2019
, Cambridge, United Kingdom, 3/09/19. https://doi.org/10.1109/ACIIW.2019.8925192

Real-time pain detection in facial expressions for health robotics. / Dai, Laduona; Broekens, Joost; Truong, Khiet Phuong.

2019 8th International Conference on Affective Computing and Intelligent Interaction Workshops and Demos (ACIIW). Piscataway, NJ : IEEE, 2019. p. 277-283.

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

TY - GEN

T1 - Real-time pain detection in facial expressions for health robotics

AU - Dai, Laduona

AU - Broekens, Joost

AU - Truong, Khiet Phuong

PY - 2019

Y1 - 2019

N2 - Automatic pain detection is an important challenge in health computing. In this paper we report on our efforts to develop a real-time, real-world pain detection system from human facial expressions. Although many studies addressed this challenge, most of them use the same dataset for training and testing. There is no cross-check with other datasets or implementation in real-time to check performance on new data. This is problematic, as evidenced in this paper, because the classifiers overtrain on dataset-specific features. This limits realtime, real-world usage. In this paper, we investigate different methods of real-time pain detection. The training data uses a combination of pain and emotion datasets, unlike other papers. The best model shows an accuracy of 88.4% on a dataset including pain and 7 non-pain emotional expressions. Results suggest that convolutional neural networks (CNN) are not the best methods in some cases as they easily overtrain if the dataset is biased. Finally we implemented our pain detection method on a humanoid robot for physiotherapy. Our work highlights the importance of cross-corpus evaluation & real-time testing, as well as the need for a well balanced and ecologically valid pain dataset.

AB - Automatic pain detection is an important challenge in health computing. In this paper we report on our efforts to develop a real-time, real-world pain detection system from human facial expressions. Although many studies addressed this challenge, most of them use the same dataset for training and testing. There is no cross-check with other datasets or implementation in real-time to check performance on new data. This is problematic, as evidenced in this paper, because the classifiers overtrain on dataset-specific features. This limits realtime, real-world usage. In this paper, we investigate different methods of real-time pain detection. The training data uses a combination of pain and emotion datasets, unlike other papers. The best model shows an accuracy of 88.4% on a dataset including pain and 7 non-pain emotional expressions. Results suggest that convolutional neural networks (CNN) are not the best methods in some cases as they easily overtrain if the dataset is biased. Finally we implemented our pain detection method on a humanoid robot for physiotherapy. Our work highlights the importance of cross-corpus evaluation & real-time testing, as well as the need for a well balanced and ecologically valid pain dataset.

KW - Pain detection

KW - Classification

KW - Generalization

KW - Cross validation

KW - Health

U2 - 10.1109/ACIIW.2019.8925192

DO - 10.1109/ACIIW.2019.8925192

M3 - Conference contribution

SN - 978-1-7281-3892-3

SP - 277

EP - 283

BT - 2019 8th International Conference on Affective Computing and Intelligent Interaction Workshops and Demos (ACIIW)

PB - IEEE

CY - Piscataway, NJ

ER -

Dai L, Broekens J, Truong KP. Real-time pain detection in facial expressions for health robotics. In 2019 8th International Conference on Affective Computing and Intelligent Interaction Workshops and Demos (ACIIW). Piscataway, NJ: IEEE. 2019. p. 277-283 https://doi.org/10.1109/ACIIW.2019.8925192