CNN Learning Strategy for Recognizing Facial Expressions
The ability to recognize facial expressions using computer vision is a crucial task that has numerous potential applications. Although deep neural networks have achieved high performance, their use in the recognition of facial expressions is still challenging. This is because different facial expres...
Main Authors: | , |
---|---|
Format: | Article |
Language: | English |
Published: |
IEEE
2023-01-01
|
Series: | IEEE Access |
Subjects: | |
Online Access: | https://ieeexplore.ieee.org/document/10177800/ |
_version_ | 1797776916572274688 |
---|---|
author | Dong-Hwan Lee Jang-Hee Yoo |
author_facet | Dong-Hwan Lee Jang-Hee Yoo |
author_sort | Dong-Hwan Lee |
collection | DOAJ |
description | The ability to recognize facial expressions using computer vision is a crucial task that has numerous potential applications. Although deep neural networks have achieved high performance, their use in the recognition of facial expressions is still challenging. This is because different facial expressions have varying degrees of similarities among themselves, and numerous variations cause diversity in the same facial images. In this study, we propose a novel divide-and-conquer-based learning strategy to improve the performance of facial expression recognition (FER). The face area in an image was detected using MobileNet, and a ResNet-18 model was employed as a backbone deep neural network for recognizing facial expressions. Subsequently, groups containing similar facial expressions were categorized by analyzing the confusion matrix, which represents the inference results of the trained ResNet-18 model, and these similar facial expression groups were then utilized to re-train the deep learning model. In the experiments, the proposed method was evaluated using two thermal (Tufts and RWTH) and two RGB (RAF and FER2013) datasets. The results demonstrate improved FER performance, with an accuracy of 97.75% for Tufts, 86.11% for RWTH, 90.81% for RAF, and 77.83% for FER2013. As such, the proposed method can accurately classify large amounts of facial expression data. |
first_indexed | 2024-03-12T22:56:40Z |
format | Article |
id | doaj.art-306370603d594ae58a00c3dc0dd939ec |
institution | Directory Open Access Journal |
issn | 2169-3536 |
language | English |
last_indexed | 2024-03-12T22:56:40Z |
publishDate | 2023-01-01 |
publisher | IEEE |
record_format | Article |
series | IEEE Access |
spelling | doaj.art-306370603d594ae58a00c3dc0dd939ec2023-07-19T23:00:32ZengIEEEIEEE Access2169-35362023-01-0111708657087210.1109/ACCESS.2023.329409910177800CNN Learning Strategy for Recognizing Facial ExpressionsDong-Hwan Lee0https://orcid.org/0009-0001-2181-9213Jang-Hee Yoo1https://orcid.org/0000-0003-0495-9211Electronics and Telecommunications Research Institute (ETRI), Daejeon, South KoreaElectronics and Telecommunications Research Institute (ETRI), Daejeon, South KoreaThe ability to recognize facial expressions using computer vision is a crucial task that has numerous potential applications. Although deep neural networks have achieved high performance, their use in the recognition of facial expressions is still challenging. This is because different facial expressions have varying degrees of similarities among themselves, and numerous variations cause diversity in the same facial images. In this study, we propose a novel divide-and-conquer-based learning strategy to improve the performance of facial expression recognition (FER). The face area in an image was detected using MobileNet, and a ResNet-18 model was employed as a backbone deep neural network for recognizing facial expressions. Subsequently, groups containing similar facial expressions were categorized by analyzing the confusion matrix, which represents the inference results of the trained ResNet-18 model, and these similar facial expression groups were then utilized to re-train the deep learning model. In the experiments, the proposed method was evaluated using two thermal (Tufts and RWTH) and two RGB (RAF and FER2013) datasets. The results demonstrate improved FER performance, with an accuracy of 97.75% for Tufts, 86.11% for RWTH, 90.81% for RAF, and 77.83% for FER2013. As such, the proposed method can accurately classify large amounts of facial expression data.https://ieeexplore.ieee.org/document/10177800/Convolutional neural networkdivide-and-conquerfacial expression recognitionlearning strategy |
spellingShingle | Dong-Hwan Lee Jang-Hee Yoo CNN Learning Strategy for Recognizing Facial Expressions IEEE Access Convolutional neural network divide-and-conquer facial expression recognition learning strategy |
title | CNN Learning Strategy for Recognizing Facial Expressions |
title_full | CNN Learning Strategy for Recognizing Facial Expressions |
title_fullStr | CNN Learning Strategy for Recognizing Facial Expressions |
title_full_unstemmed | CNN Learning Strategy for Recognizing Facial Expressions |
title_short | CNN Learning Strategy for Recognizing Facial Expressions |
title_sort | cnn learning strategy for recognizing facial expressions |
topic | Convolutional neural network divide-and-conquer facial expression recognition learning strategy |
url | https://ieeexplore.ieee.org/document/10177800/ |
work_keys_str_mv | AT donghwanlee cnnlearningstrategyforrecognizingfacialexpressions AT jangheeyoo cnnlearningstrategyforrecognizingfacialexpressions |