Device Position-Independent Human Activity Recognition with Wearable Sensors Using Deep Neural Networks
Human activity recognition (HAR) identifies people’s motions and actions in daily life. HAR research has grown with the popularity of internet-connected, wearable sensors that capture human movement data to detect activities. Recent deep learning advances have enabled more HAR research and applicati...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2024-03-01
|
| Series: | Applied Sciences |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2076-3417/14/5/2107 |
| _version_ | 1797264811476647936 |
|---|---|
| author | Sakorn Mekruksavanich Anuchit Jitpattanakul |
| author_facet | Sakorn Mekruksavanich Anuchit Jitpattanakul |
| author_sort | Sakorn Mekruksavanich |
| collection | DOAJ |
| description | Human activity recognition (HAR) identifies people’s motions and actions in daily life. HAR research has grown with the popularity of internet-connected, wearable sensors that capture human movement data to detect activities. Recent deep learning advances have enabled more HAR research and applications using data from wearable devices. However, prior HAR research often focused on a few sensor locations on the body. Recognizing real-world activities poses challenges when device positioning is uncontrolled or initial user training data are unavailable. This research analyzes the feasibility of deep learning models for both position-dependent and position-independent HAR. We introduce an advanced residual deep learning model called Att-ResBiGRU, which excels at accurate position-dependent HAR and delivers excellent performance for position-independent HAR. We evaluate this model using three public HAR datasets: Opportunity, PAMAP2, and REALWORLD16. Comparisons are made to previously published deep learning architectures for addressing HAR challenges. The proposed Att-ResBiGRU model outperforms existing techniques in accuracy, cross-entropy loss, and F1-score across all three datasets. We assess the model using k-fold cross-validation. The Att-ResBiGRU achieves F1-scores of 86.69%, 96.23%, and 96.44% on the PAMAP2, REALWORLD16, and Opportunity datasets, surpassing state-of-the-art models across all datasets. Our experiments and analysis demonstrate the exceptional performance of the Att-ResBiGRU model for HAR applications. |
| first_indexed | 2024-04-25T00:34:50Z |
| format | Article |
| id | doaj.art-dc2b37ee84ae4d2b89cfddf7c086e0f9 |
| institution | Directory Open Access Journal |
| issn | 2076-3417 |
| language | English |
| last_indexed | 2024-04-25T00:34:50Z |
| publishDate | 2024-03-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Applied Sciences |
| spelling | doaj.art-dc2b37ee84ae4d2b89cfddf7c086e0f92024-03-12T16:40:08ZengMDPI AGApplied Sciences2076-34172024-03-01145210710.3390/app14052107Device Position-Independent Human Activity Recognition with Wearable Sensors Using Deep Neural NetworksSakorn Mekruksavanich0Anuchit Jitpattanakul1Department of Computer Engineering, School of Information and Communication Technology, University of Phayao, Phayao 56000, ThailandDepartment of Mathematics, Faculty of Applied Science, King Mongkut’s University of Technology North Bangkok, Bangkok 10800, ThailandHuman activity recognition (HAR) identifies people’s motions and actions in daily life. HAR research has grown with the popularity of internet-connected, wearable sensors that capture human movement data to detect activities. Recent deep learning advances have enabled more HAR research and applications using data from wearable devices. However, prior HAR research often focused on a few sensor locations on the body. Recognizing real-world activities poses challenges when device positioning is uncontrolled or initial user training data are unavailable. This research analyzes the feasibility of deep learning models for both position-dependent and position-independent HAR. We introduce an advanced residual deep learning model called Att-ResBiGRU, which excels at accurate position-dependent HAR and delivers excellent performance for position-independent HAR. We evaluate this model using three public HAR datasets: Opportunity, PAMAP2, and REALWORLD16. Comparisons are made to previously published deep learning architectures for addressing HAR challenges. The proposed Att-ResBiGRU model outperforms existing techniques in accuracy, cross-entropy loss, and F1-score across all three datasets. We assess the model using k-fold cross-validation. The Att-ResBiGRU achieves F1-scores of 86.69%, 96.23%, and 96.44% on the PAMAP2, REALWORLD16, and Opportunity datasets, surpassing state-of-the-art models across all datasets. Our experiments and analysis demonstrate the exceptional performance of the Att-ResBiGRU model for HAR applications.https://www.mdpi.com/2076-3417/14/5/2107human activity recognitionposition-independent sensingwearable sensorsdeep learningresidual neural network |
| spellingShingle | Sakorn Mekruksavanich Anuchit Jitpattanakul Device Position-Independent Human Activity Recognition with Wearable Sensors Using Deep Neural Networks Applied Sciences human activity recognition position-independent sensing wearable sensors deep learning residual neural network |
| title | Device Position-Independent Human Activity Recognition with Wearable Sensors Using Deep Neural Networks |
| title_full | Device Position-Independent Human Activity Recognition with Wearable Sensors Using Deep Neural Networks |
| title_fullStr | Device Position-Independent Human Activity Recognition with Wearable Sensors Using Deep Neural Networks |
| title_full_unstemmed | Device Position-Independent Human Activity Recognition with Wearable Sensors Using Deep Neural Networks |
| title_short | Device Position-Independent Human Activity Recognition with Wearable Sensors Using Deep Neural Networks |
| title_sort | device position independent human activity recognition with wearable sensors using deep neural networks |
| topic | human activity recognition position-independent sensing wearable sensors deep learning residual neural network |
| url | https://www.mdpi.com/2076-3417/14/5/2107 |
| work_keys_str_mv | AT sakornmekruksavanich devicepositionindependenthumanactivityrecognitionwithwearablesensorsusingdeepneuralnetworks AT anuchitjitpattanakul devicepositionindependenthumanactivityrecognitionwithwearablesensorsusingdeepneuralnetworks |