Home > Research > Publications & Outputs > Joint Learning of Temporal Models to Handle Imb...

Research

Links

Text available via DOI:

Keywords

View graph of relations

Joint Learning of Temporal Models to Handle Imbalanced Data for Human Activity Recognition

Research output: Contribution to Journal/Magazine › Journal article › peer-review

Published

Standard

Joint Learning of Temporal Models to Handle Imbalanced Data for Human Activity Recognition. / Hamad, Rebeen Ali; Yang, Longzhi; Woo, Wai Lok et al.
In: Applied Sciences, Vol. 10, No. 15, 5293, 01.08.2020.

Research output: Contribution to Journal/Magazine › Journal article › peer-review

Harvard

Hamad, RA, Yang, L, Woo, WL & Wei, B 2020, 'Joint Learning of Temporal Models to Handle Imbalanced Data for Human Activity Recognition', Applied Sciences, vol. 10, no. 15, 5293. https://doi.org/10.3390/app10155293

APA

Hamad, R. A., Yang, L., Woo, W. L., & Wei, B. (2020). Joint Learning of Temporal Models to Handle Imbalanced Data for Human Activity Recognition. Applied Sciences, 10(15), Article 5293. https://doi.org/10.3390/app10155293

Vancouver

Hamad RA, Yang L, Woo WL, Wei B. Joint Learning of Temporal Models to Handle Imbalanced Data for Human Activity Recognition. Applied Sciences. 2020 Aug 1;10(15):5293. Epub 2020 Jul 30. doi: 10.3390/app10155293

Author

Hamad, Rebeen Ali ; Yang, Longzhi ; Woo, Wai Lok et al. / Joint Learning of Temporal Models to Handle Imbalanced Data for Human Activity Recognition. In: Applied Sciences. 2020 ; Vol. 10, No. 15.

Bibtex

@article{0561b4590c5b40279c39642cac297cb5,
title = "Joint Learning of Temporal Models to Handle Imbalanced Data for Human Activity Recognition",
abstract = "Human activity recognition has become essential to a wide range of applications, such as smart home monitoring, health-care, surveillance. However, it is challenging to deliver a sufficiently robust human activity recognition system from raw sensor data with noise in a smart environment setting. Moreover, imbalanced human activity datasets with less frequent activities create extra challenges for accurate activity recognition. Deep learning algorithms have achieved promisingresults on balanced datasets, but their performance on imbalanced datasets without explicit algorithm design cannot be promised. Therefore, we aim to realise an activity recognition system using multi-modal sensors to address the issue of class imbalance in deep learning and improve recognition accuracy. This paper proposes a joint diverse temporal learning framework using Long Short Term Memory and one-dimensional Convolutional Neural Network models to improve human activity recognition, especially for less represented activities. We extensively evaluate the proposed method for Activities of Daily Living recognition using binary sensors dataset. A comparative study on five smart home datasets demonstrate that our proposed approach outperforms the existing individual temporal models and their hybridization. Furthermore, this is particularly the case for minority classes in addition to reasonable improvement on the majority classes of human activities.",
keywords = "Activity recognition, Smart home, Imbalanced class, Joint learning, Temporal models",
author = "Hamad, {Rebeen Ali} and Longzhi Yang and Woo, {Wai Lok} and Bo Wei",
year = "2020",
month = aug,
day = "1",
doi = "10.3390/app10155293",
language = "English",
volume = "10",
journal = "Applied Sciences",
issn = "2076-3417",
publisher = "Multidisciplinary Digital Publishing Institute (MDPI)",
number = "15",

}

RIS

TY - JOUR

T1 - Joint Learning of Temporal Models to Handle Imbalanced Data for Human Activity Recognition

AU - Hamad, Rebeen Ali

AU - Yang, Longzhi

AU - Woo, Wai Lok

AU - Wei, Bo

PY - 2020/8/1

Y1 - 2020/8/1

N2 - Human activity recognition has become essential to a wide range of applications, such as smart home monitoring, health-care, surveillance. However, it is challenging to deliver a sufficiently robust human activity recognition system from raw sensor data with noise in a smart environment setting. Moreover, imbalanced human activity datasets with less frequent activities create extra challenges for accurate activity recognition. Deep learning algorithms have achieved promisingresults on balanced datasets, but their performance on imbalanced datasets without explicit algorithm design cannot be promised. Therefore, we aim to realise an activity recognition system using multi-modal sensors to address the issue of class imbalance in deep learning and improve recognition accuracy. This paper proposes a joint diverse temporal learning framework using Long Short Term Memory and one-dimensional Convolutional Neural Network models to improve human activity recognition, especially for less represented activities. We extensively evaluate the proposed method for Activities of Daily Living recognition using binary sensors dataset. A comparative study on five smart home datasets demonstrate that our proposed approach outperforms the existing individual temporal models and their hybridization. Furthermore, this is particularly the case for minority classes in addition to reasonable improvement on the majority classes of human activities.

AB - Human activity recognition has become essential to a wide range of applications, such as smart home monitoring, health-care, surveillance. However, it is challenging to deliver a sufficiently robust human activity recognition system from raw sensor data with noise in a smart environment setting. Moreover, imbalanced human activity datasets with less frequent activities create extra challenges for accurate activity recognition. Deep learning algorithms have achieved promisingresults on balanced datasets, but their performance on imbalanced datasets without explicit algorithm design cannot be promised. Therefore, we aim to realise an activity recognition system using multi-modal sensors to address the issue of class imbalance in deep learning and improve recognition accuracy. This paper proposes a joint diverse temporal learning framework using Long Short Term Memory and one-dimensional Convolutional Neural Network models to improve human activity recognition, especially for less represented activities. We extensively evaluate the proposed method for Activities of Daily Living recognition using binary sensors dataset. A comparative study on five smart home datasets demonstrate that our proposed approach outperforms the existing individual temporal models and their hybridization. Furthermore, this is particularly the case for minority classes in addition to reasonable improvement on the majority classes of human activities.

KW - Activity recognition

KW - Smart home

KW - Imbalanced class

KW - Joint learning

KW - Temporal models

U2 - 10.3390/app10155293

DO - 10.3390/app10155293

M3 - Journal article

VL - 10

JO - Applied Sciences

JF - Applied Sciences

SN - 2076-3417

IS - 15

M1 - 5293

ER -