ResNet-SE: Channel Attention-Based Deep Residual Network for Complex Activity Recognition Using Wrist-Worn Wearable Sensors
Smart mobile devices are being widely used to identify and track human behaviors in simple and complex daily activities. The evolution of wearable sensing technologies pertaining to wellness, living surveillance, and fitness tracking is based on the accurate analysis of people’s behavior...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2022-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/9771436/ |
| _version_ | 1818204580812423168 |
|---|---|
| author | Sakorn Mekruksavanich Anuchit Jitpattanakul Kanokwan Sitthithakerngkiet Phichai Youplao Preecha Yupapin |
| author_facet | Sakorn Mekruksavanich Anuchit Jitpattanakul Kanokwan Sitthithakerngkiet Phichai Youplao Preecha Yupapin |
| author_sort | Sakorn Mekruksavanich |
| collection | DOAJ |
| description | Smart mobile devices are being widely used to identify and track human behaviors in simple and complex daily activities. The evolution of wearable sensing technologies pertaining to wellness, living surveillance, and fitness tracking is based on the accurate analysis of people’s behavior from the data acquired through different sensors embedded in smart devices, especially wrist-worn wearable technologies such as smartwatches. Many deep learning techniques have been developed to realize human activity recognition (HAR), with simple daily activities being focused on. However, several challenges remain to be addressed in complex HAR research involving specific human behaviors in different contexts. To address the problems pertaining to complex HAR, a deep neural network composed of convolutional layers and residual networks was developed in this work. Additional attention was incorporated in the system by using a squeeze-and-excite mechanism. The model effectiveness was investigated considering three publicly available datasets, (WISDM-HARB, UT-Smoke, and UT-Complex). The proposed network achieved overall accuracies of 94.91%, 98.75%, and 97.73% over WISDM-HARB, UT-Smoke, and UT-Complex, respectively. The results showed that deep residual networks are more durable and superior at activity recognition than the existing models. |
| first_indexed | 2024-12-12T03:43:30Z |
| format | Article |
| id | doaj.art-ffdd7542ce9d4f1d90bb4fba241bdcc4 |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-12-12T03:43:30Z |
| publishDate | 2022-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-ffdd7542ce9d4f1d90bb4fba241bdcc42022-12-22T00:39:37ZengIEEEIEEE Access2169-35362022-01-0110511425115410.1109/ACCESS.2022.31741249771436ResNet-SE: Channel Attention-Based Deep Residual Network for Complex Activity Recognition Using Wrist-Worn Wearable SensorsSakorn Mekruksavanich0https://orcid.org/0000-0002-3735-4262Anuchit Jitpattanakul1https://orcid.org/0000-0002-5249-2786Kanokwan Sitthithakerngkiet2https://orcid.org/0000-0002-8496-7803Phichai Youplao3https://orcid.org/0000-0002-4280-0482Preecha Yupapin4https://orcid.org/0000-0002-5257-4351Department of Computer Engineering, School of Information and Communication Technology, University of Phayao, Phayao, ThailandDepartment of Mathematics, Faculty of Applied Science, Intelligent and Nonlinear Dynamic Innovations Research Center, King Mongkut’s University of Technology North Bangkok, Bangkok, ThailandDepartment of Mathematics, Faculty of Applied Science, Intelligent and Nonlinear Dynamic Innovations Research Center, King Mongkut’s University of Technology North Bangkok, Bangkok, ThailandDepartment of Electrical Engineering, Faculty of Industry and Technology, Rajamangala University of Technology Isan Sakon Nakhon Campus, Sakon Nakhon, ThailandDepartment of Electrical Technology, Faculty of Industrial Technology, Institute of Vocational Education Northeastern 2, Sakon Nakhon, ThailandSmart mobile devices are being widely used to identify and track human behaviors in simple and complex daily activities. The evolution of wearable sensing technologies pertaining to wellness, living surveillance, and fitness tracking is based on the accurate analysis of people’s behavior from the data acquired through different sensors embedded in smart devices, especially wrist-worn wearable technologies such as smartwatches. Many deep learning techniques have been developed to realize human activity recognition (HAR), with simple daily activities being focused on. However, several challenges remain to be addressed in complex HAR research involving specific human behaviors in different contexts. To address the problems pertaining to complex HAR, a deep neural network composed of convolutional layers and residual networks was developed in this work. Additional attention was incorporated in the system by using a squeeze-and-excite mechanism. The model effectiveness was investigated considering three publicly available datasets, (WISDM-HARB, UT-Smoke, and UT-Complex). The proposed network achieved overall accuracies of 94.91%, 98.75%, and 97.73% over WISDM-HARB, UT-Smoke, and UT-Complex, respectively. The results showed that deep residual networks are more durable and superior at activity recognition than the existing models.https://ieeexplore.ieee.org/document/9771436/Wrist-worn wearable sensordeep learningdeep residual networkattention mechanismcomplex activity recognition |
| spellingShingle | Sakorn Mekruksavanich Anuchit Jitpattanakul Kanokwan Sitthithakerngkiet Phichai Youplao Preecha Yupapin ResNet-SE: Channel Attention-Based Deep Residual Network for Complex Activity Recognition Using Wrist-Worn Wearable Sensors IEEE Access Wrist-worn wearable sensor deep learning deep residual network attention mechanism complex activity recognition |
| title | ResNet-SE: Channel Attention-Based Deep Residual Network for Complex Activity Recognition Using Wrist-Worn Wearable Sensors |
| title_full | ResNet-SE: Channel Attention-Based Deep Residual Network for Complex Activity Recognition Using Wrist-Worn Wearable Sensors |
| title_fullStr | ResNet-SE: Channel Attention-Based Deep Residual Network for Complex Activity Recognition Using Wrist-Worn Wearable Sensors |
| title_full_unstemmed | ResNet-SE: Channel Attention-Based Deep Residual Network for Complex Activity Recognition Using Wrist-Worn Wearable Sensors |
| title_short | ResNet-SE: Channel Attention-Based Deep Residual Network for Complex Activity Recognition Using Wrist-Worn Wearable Sensors |
| title_sort | resnet se channel attention based deep residual network for complex activity recognition using wrist worn wearable sensors |
| topic | Wrist-worn wearable sensor deep learning deep residual network attention mechanism complex activity recognition |
| url | https://ieeexplore.ieee.org/document/9771436/ |
| work_keys_str_mv | AT sakornmekruksavanich resnetsechannelattentionbaseddeepresidualnetworkforcomplexactivityrecognitionusingwristwornwearablesensors AT anuchitjitpattanakul resnetsechannelattentionbaseddeepresidualnetworkforcomplexactivityrecognitionusingwristwornwearablesensors AT kanokwansitthithakerngkiet resnetsechannelattentionbaseddeepresidualnetworkforcomplexactivityrecognitionusingwristwornwearablesensors AT phichaiyouplao resnetsechannelattentionbaseddeepresidualnetworkforcomplexactivityrecognitionusingwristwornwearablesensors AT preechayupapin resnetsechannelattentionbaseddeepresidualnetworkforcomplexactivityrecognitionusingwristwornwearablesensors |