Detailed human activity recognition based on multiple HMM

Mariana Abreu, Marília Barandas, Ricardo Leonardo, Hugo Gamboa

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

A wide array of activities is performed by humans, everyday. In healthcare, precocious detection of movement changes in daily activities and their monitoring, are important contributors to assess the patient general wellbeing. Several previous studies are successful in activity recognition, but few of them provide a meticulous discrimination. Hereby, we created a novel framework specialized in detailed human activities, where signals from four sensors were used: accelerometer, gyroscope, magnetometer and microphone. A new dataset was created, with 10 complex activities, suchlike opening a door, brushing the teeth and typing on the keyboard. The classifier was based on multiple hidden Markov models, one per activity. The developed solution was evaluated in the offline context, where it achieved an accuracy of 84±4.8%. It also showed a solid performance in other performed tests, where it was tested with different detailed activities, and in simulations of real time recognition. This solution can be applied in elderly monitoring to access their well-being and also in the early detection of degenerative diseases.

Original languageEnglish
Title of host publicationBIOSIGNALS 2019 - 12th International Conference on Bio-Inspired Systems and Signal Processing, Proceedings; Part of 12th International Joint Conference on Biomedical Engineering Systems and Technologies, BIOSTEC 2019
EditorsFelix Putze, Ana Fred, Hugo Gamboa
PublisherSciTePress
Pages171-178
Number of pages8
ISBN (Electronic)9789897583537
Publication statusPublished - 1 Jan 2019
Event12th International Conference on Bio-Inspired Systems and Signal Processing, BIOSIGNALS 2019 - Part of 12th International Joint Conference on Biomedical Engineering Systems and Technologies, BIOSTEC 2019 - Prague, Czech Republic
Duration: 22 Feb 201924 Feb 2019

Conference

Conference12th International Conference on Bio-Inspired Systems and Signal Processing, BIOSIGNALS 2019 - Part of 12th International Joint Conference on Biomedical Engineering Systems and Technologies, BIOSTEC 2019
CountryCzech Republic
CityPrague
Period22/02/1924/02/19

Fingerprint

Monitoring
Gyroscopes
Magnetometers
Hidden Markov models
Microphones
Accelerometers
Classifiers
Sensors

Keywords

  • Feature selection
  • Gesture recognition
  • Hidden Markov models
  • Human activity recognition
  • Smartphone sensors

Cite this

Abreu, M., Barandas, M., Leonardo, R., & Gamboa, H. (2019). Detailed human activity recognition based on multiple HMM. In F. Putze, A. Fred, & H. Gamboa (Eds.), BIOSIGNALS 2019 - 12th International Conference on Bio-Inspired Systems and Signal Processing, Proceedings; Part of 12th International Joint Conference on Biomedical Engineering Systems and Technologies, BIOSTEC 2019 (pp. 171-178). SciTePress.
Abreu, Mariana ; Barandas, Marília ; Leonardo, Ricardo ; Gamboa, Hugo. / Detailed human activity recognition based on multiple HMM. BIOSIGNALS 2019 - 12th International Conference on Bio-Inspired Systems and Signal Processing, Proceedings; Part of 12th International Joint Conference on Biomedical Engineering Systems and Technologies, BIOSTEC 2019. editor / Felix Putze ; Ana Fred ; Hugo Gamboa. SciTePress, 2019. pp. 171-178
@inproceedings{9a8ec30b984a409a80f4039cf7f0ba13,
title = "Detailed human activity recognition based on multiple HMM",
abstract = "A wide array of activities is performed by humans, everyday. In healthcare, precocious detection of movement changes in daily activities and their monitoring, are important contributors to assess the patient general wellbeing. Several previous studies are successful in activity recognition, but few of them provide a meticulous discrimination. Hereby, we created a novel framework specialized in detailed human activities, where signals from four sensors were used: accelerometer, gyroscope, magnetometer and microphone. A new dataset was created, with 10 complex activities, suchlike opening a door, brushing the teeth and typing on the keyboard. The classifier was based on multiple hidden Markov models, one per activity. The developed solution was evaluated in the offline context, where it achieved an accuracy of 84±4.8{\%}. It also showed a solid performance in other performed tests, where it was tested with different detailed activities, and in simulations of real time recognition. This solution can be applied in elderly monitoring to access their well-being and also in the early detection of degenerative diseases.",
keywords = "Feature selection, Gesture recognition, Hidden Markov models, Human activity recognition, Smartphone sensors",
author = "Mariana Abreu and Mar{\'i}lia Barandas and Ricardo Leonardo and Hugo Gamboa",
note = "This work was supported by North Portugal Regional Operational Programme (NORTE 2020), Portugal 2020 and the European Regional Development Fund (ERDF) from European Union through the project Symbiotic technology for societal efficiency gains: Deus ex Machina (DEM) [NORTE-01-0145-FEDER-000026]",
year = "2019",
month = "1",
day = "1",
language = "English",
pages = "171--178",
editor = "Felix Putze and Ana Fred and Hugo Gamboa",
booktitle = "BIOSIGNALS 2019 - 12th International Conference on Bio-Inspired Systems and Signal Processing, Proceedings; Part of 12th International Joint Conference on Biomedical Engineering Systems and Technologies, BIOSTEC 2019",
publisher = "SciTePress",

}

Abreu, M, Barandas, M, Leonardo, R & Gamboa, H 2019, Detailed human activity recognition based on multiple HMM. in F Putze, A Fred & H Gamboa (eds), BIOSIGNALS 2019 - 12th International Conference on Bio-Inspired Systems and Signal Processing, Proceedings; Part of 12th International Joint Conference on Biomedical Engineering Systems and Technologies, BIOSTEC 2019. SciTePress, pp. 171-178, 12th International Conference on Bio-Inspired Systems and Signal Processing, BIOSIGNALS 2019 - Part of 12th International Joint Conference on Biomedical Engineering Systems and Technologies, BIOSTEC 2019, Prague, Czech Republic, 22/02/19.

Detailed human activity recognition based on multiple HMM. / Abreu, Mariana; Barandas, Marília; Leonardo, Ricardo; Gamboa, Hugo.

BIOSIGNALS 2019 - 12th International Conference on Bio-Inspired Systems and Signal Processing, Proceedings; Part of 12th International Joint Conference on Biomedical Engineering Systems and Technologies, BIOSTEC 2019. ed. / Felix Putze; Ana Fred; Hugo Gamboa. SciTePress, 2019. p. 171-178.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Detailed human activity recognition based on multiple HMM

AU - Abreu, Mariana

AU - Barandas, Marília

AU - Leonardo, Ricardo

AU - Gamboa, Hugo

N1 - This work was supported by North Portugal Regional Operational Programme (NORTE 2020), Portugal 2020 and the European Regional Development Fund (ERDF) from European Union through the project Symbiotic technology for societal efficiency gains: Deus ex Machina (DEM) [NORTE-01-0145-FEDER-000026]

PY - 2019/1/1

Y1 - 2019/1/1

N2 - A wide array of activities is performed by humans, everyday. In healthcare, precocious detection of movement changes in daily activities and their monitoring, are important contributors to assess the patient general wellbeing. Several previous studies are successful in activity recognition, but few of them provide a meticulous discrimination. Hereby, we created a novel framework specialized in detailed human activities, where signals from four sensors were used: accelerometer, gyroscope, magnetometer and microphone. A new dataset was created, with 10 complex activities, suchlike opening a door, brushing the teeth and typing on the keyboard. The classifier was based on multiple hidden Markov models, one per activity. The developed solution was evaluated in the offline context, where it achieved an accuracy of 84±4.8%. It also showed a solid performance in other performed tests, where it was tested with different detailed activities, and in simulations of real time recognition. This solution can be applied in elderly monitoring to access their well-being and also in the early detection of degenerative diseases.

AB - A wide array of activities is performed by humans, everyday. In healthcare, precocious detection of movement changes in daily activities and their monitoring, are important contributors to assess the patient general wellbeing. Several previous studies are successful in activity recognition, but few of them provide a meticulous discrimination. Hereby, we created a novel framework specialized in detailed human activities, where signals from four sensors were used: accelerometer, gyroscope, magnetometer and microphone. A new dataset was created, with 10 complex activities, suchlike opening a door, brushing the teeth and typing on the keyboard. The classifier was based on multiple hidden Markov models, one per activity. The developed solution was evaluated in the offline context, where it achieved an accuracy of 84±4.8%. It also showed a solid performance in other performed tests, where it was tested with different detailed activities, and in simulations of real time recognition. This solution can be applied in elderly monitoring to access their well-being and also in the early detection of degenerative diseases.

KW - Feature selection

KW - Gesture recognition

KW - Hidden Markov models

KW - Human activity recognition

KW - Smartphone sensors

UR - http://www.scopus.com/inward/record.url?scp=85064731374&partnerID=8YFLogxK

M3 - Conference contribution

SP - 171

EP - 178

BT - BIOSIGNALS 2019 - 12th International Conference on Bio-Inspired Systems and Signal Processing, Proceedings; Part of 12th International Joint Conference on Biomedical Engineering Systems and Technologies, BIOSTEC 2019

A2 - Putze, Felix

A2 - Fred, Ana

A2 - Gamboa, Hugo

PB - SciTePress

ER -

Abreu M, Barandas M, Leonardo R, Gamboa H. Detailed human activity recognition based on multiple HMM. In Putze F, Fred A, Gamboa H, editors, BIOSIGNALS 2019 - 12th International Conference on Bio-Inspired Systems and Signal Processing, Proceedings; Part of 12th International Joint Conference on Biomedical Engineering Systems and Technologies, BIOSTEC 2019. SciTePress. 2019. p. 171-178