A Systematic Exploration of Deep Neural Networks for EDA-Based Emotion Recognition
Subject-independent emotion recognition based on physiological signals has become a research hotspot. Previous research has proved that electrodermal activity (EDA) signals are an effective data resource for emotion recognition. Benefiting from their great representation ability, an increasing numbe...
Main Authors: | , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2020-04-01
|
Series: | Information |
Subjects: | |
Online Access: | https://www.mdpi.com/2078-2489/11/4/212 |
id |
doaj-616db4eb1ca44e20a890e765284949b1 |
---|---|
record_format |
Article |
spelling |
doaj-616db4eb1ca44e20a890e765284949b12020-11-25T03:16:36ZengMDPI AGInformation2078-24892020-04-011121221210.3390/info11040212A Systematic Exploration of Deep Neural Networks for EDA-Based Emotion RecognitionDian Yu0Shouqian Sun1College of Computer Science and Technology, Zhejiang University, Hangzhou 310037, ChinaCollege of Computer Science and Technology, Zhejiang University, Hangzhou 310037, ChinaSubject-independent emotion recognition based on physiological signals has become a research hotspot. Previous research has proved that electrodermal activity (EDA) signals are an effective data resource for emotion recognition. Benefiting from their great representation ability, an increasing number of deep neural networks have been applied for emotion recognition, and they can be classified as a Convolutional Neural Network (CNN), a Recurrent Neural Network (RNN), or a combination of these (CNN+RNN). However, there has been no systematic research on the predictive power and configurations of different deep neural networks in this task. In this work, we systematically explore the configurations and performances of three adapted deep neural networks: ResNet, LSTM, and hybrid ResNet-LSTM. Our experiments use the subject-independent method to evaluate the three-class classification on the MAHNOB dataset. The results prove that the CNN model (ResNet) reaches a better accuracy and F1 score than the RNN model (LSTM) and the CNN+RNN model (hybrid ResNet-LSTM). Extensive comparisons also reveal that our three deep neural networks with EDA data outperform previous models with handcraft features on emotion recognition, which proves the great potential of the end-to-end DNN method.https://www.mdpi.com/2078-2489/11/4/212emotion recognitionelectrodermal activitydeep neural network |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Dian Yu Shouqian Sun |
spellingShingle |
Dian Yu Shouqian Sun A Systematic Exploration of Deep Neural Networks for EDA-Based Emotion Recognition Information emotion recognition electrodermal activity deep neural network |
author_facet |
Dian Yu Shouqian Sun |
author_sort |
Dian Yu |
title |
A Systematic Exploration of Deep Neural Networks for EDA-Based Emotion Recognition |
title_short |
A Systematic Exploration of Deep Neural Networks for EDA-Based Emotion Recognition |
title_full |
A Systematic Exploration of Deep Neural Networks for EDA-Based Emotion Recognition |
title_fullStr |
A Systematic Exploration of Deep Neural Networks for EDA-Based Emotion Recognition |
title_full_unstemmed |
A Systematic Exploration of Deep Neural Networks for EDA-Based Emotion Recognition |
title_sort |
systematic exploration of deep neural networks for eda-based emotion recognition |
publisher |
MDPI AG |
series |
Information |
issn |
2078-2489 |
publishDate |
2020-04-01 |
description |
Subject-independent emotion recognition based on physiological signals has become a research hotspot. Previous research has proved that electrodermal activity (EDA) signals are an effective data resource for emotion recognition. Benefiting from their great representation ability, an increasing number of deep neural networks have been applied for emotion recognition, and they can be classified as a Convolutional Neural Network (CNN), a Recurrent Neural Network (RNN), or a combination of these (CNN+RNN). However, there has been no systematic research on the predictive power and configurations of different deep neural networks in this task. In this work, we systematically explore the configurations and performances of three adapted deep neural networks: ResNet, LSTM, and hybrid ResNet-LSTM. Our experiments use the subject-independent method to evaluate the three-class classification on the MAHNOB dataset. The results prove that the CNN model (ResNet) reaches a better accuracy and F1 score than the RNN model (LSTM) and the CNN+RNN model (hybrid ResNet-LSTM). Extensive comparisons also reveal that our three deep neural networks with EDA data outperform previous models with handcraft features on emotion recognition, which proves the great potential of the end-to-end DNN method. |
topic |
emotion recognition electrodermal activity deep neural network |
url |
https://www.mdpi.com/2078-2489/11/4/212 |
work_keys_str_mv |
AT dianyu asystematicexplorationofdeepneuralnetworksforedabasedemotionrecognition AT shouqiansun asystematicexplorationofdeepneuralnetworksforedabasedemotionrecognition AT dianyu systematicexplorationofdeepneuralnetworksforedabasedemotionrecognition AT shouqiansun systematicexplorationofdeepneuralnetworksforedabasedemotionrecognition |
_version_ |
1724635229945069568 |