Pre-Impact and Impact Fall Detection Based on a Multimodal Sensor Using a Deep Residual Network
Falls are the contributing factor to both fatal and nonfatal injuries in the elderly. Therefore, pre-impact fall detection, which identifies a fall before the body collides with the floor, would be essential. Recently, researchers have turned their attention from post-impact fall detection to pre-im...
Saved in:
Main Author: | |
---|---|
Other Authors: | |
Format: | Article |
Published: |
2023
|
Subjects: | |
Online Access: | https://repository.li.mahidol.ac.th/handle/123456789/81793 |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Institution: | Mahidol University |
id |
th-mahidol.81793 |
---|---|
record_format |
dspace |
spelling |
th-mahidol.817932023-05-19T14:39:51Z Pre-Impact and Impact Fall Detection Based on a Multimodal Sensor Using a Deep Residual Network Hnoohom N. Mahidol University Computer Science Falls are the contributing factor to both fatal and nonfatal injuries in the elderly. Therefore, pre-impact fall detection, which identifies a fall before the body collides with the floor, would be essential. Recently, researchers have turned their attention from post-impact fall detection to pre-impact fall detection. Pre-impact fall detection solutions typically use either a threshold-based or machine learning-based approach, although the threshold value would be difficult to accurately determine in threshold-based methods. Moreover, while additional features could sometimes assist in categorizing falls and non-falls more precisely, the esti-mated determination of the significant features would be too time-intensive, thus using a significant portion of the algorithm’s operating time. In this work, we developed a deep residual network with aggregation transformation called FDSNeXt for a pre-impact fall detection approach employing wearable inertial sensors. The proposed network was introduced to address the limitations of feature extraction, threshold definition, and algorithm complexity. After training on a large-scale motion dataset, the KFall dataset, and straightforward evaluation with standard metrics, the proposed approach identified pre-impact and impact falls with high accuracy of 91.87 and 92.52%, respectively. In addition, we have inves-tigated fall detection’s performances of three state-of-the-art deep learning models such as a convolutional neural network (CNN), a long short-term memory neural network (LSTM), and a hybrid model (CNN-LSTM). The experimental results showed that the proposed FDSNeXt model outperformed these deep learning models (CNN, LSTM, and CNN-LSTM) with significant improvements. 2023-05-19T07:39:51Z 2023-05-19T07:39:51Z 2023-01-01 Article Intelligent Automation and Soft Computing Vol.36 No.3 (2023) , 3371-3385 10.32604/iasc.2023.036551 2326005X 10798587 2-s2.0-85150796503 https://repository.li.mahidol.ac.th/handle/123456789/81793 SCOPUS |
institution |
Mahidol University |
building |
Mahidol University Library |
continent |
Asia |
country |
Thailand Thailand |
content_provider |
Mahidol University Library |
collection |
Mahidol University Institutional Repository |
topic |
Computer Science |
spellingShingle |
Computer Science Hnoohom N. Pre-Impact and Impact Fall Detection Based on a Multimodal Sensor Using a Deep Residual Network |
description |
Falls are the contributing factor to both fatal and nonfatal injuries in the elderly. Therefore, pre-impact fall detection, which identifies a fall before the body collides with the floor, would be essential. Recently, researchers have turned their attention from post-impact fall detection to pre-impact fall detection. Pre-impact fall detection solutions typically use either a threshold-based or machine learning-based approach, although the threshold value would be difficult to accurately determine in threshold-based methods. Moreover, while additional features could sometimes assist in categorizing falls and non-falls more precisely, the esti-mated determination of the significant features would be too time-intensive, thus using a significant portion of the algorithm’s operating time. In this work, we developed a deep residual network with aggregation transformation called FDSNeXt for a pre-impact fall detection approach employing wearable inertial sensors. The proposed network was introduced to address the limitations of feature extraction, threshold definition, and algorithm complexity. After training on a large-scale motion dataset, the KFall dataset, and straightforward evaluation with standard metrics, the proposed approach identified pre-impact and impact falls with high accuracy of 91.87 and 92.52%, respectively. In addition, we have inves-tigated fall detection’s performances of three state-of-the-art deep learning models such as a convolutional neural network (CNN), a long short-term memory neural network (LSTM), and a hybrid model (CNN-LSTM). The experimental results showed that the proposed FDSNeXt model outperformed these deep learning models (CNN, LSTM, and CNN-LSTM) with significant improvements. |
author2 |
Mahidol University |
author_facet |
Mahidol University Hnoohom N. |
format |
Article |
author |
Hnoohom N. |
author_sort |
Hnoohom N. |
title |
Pre-Impact and Impact Fall Detection Based on a Multimodal Sensor Using a Deep Residual Network |
title_short |
Pre-Impact and Impact Fall Detection Based on a Multimodal Sensor Using a Deep Residual Network |
title_full |
Pre-Impact and Impact Fall Detection Based on a Multimodal Sensor Using a Deep Residual Network |
title_fullStr |
Pre-Impact and Impact Fall Detection Based on a Multimodal Sensor Using a Deep Residual Network |
title_full_unstemmed |
Pre-Impact and Impact Fall Detection Based on a Multimodal Sensor Using a Deep Residual Network |
title_sort |
pre-impact and impact fall detection based on a multimodal sensor using a deep residual network |
publishDate |
2023 |
url |
https://repository.li.mahidol.ac.th/handle/123456789/81793 |
_version_ |
1781414470801686528 |