Skip to Main Content
 

Global Search Box

 
 
 
 

Files

File List


Thesis_final (4).pdf (37.72 MB)

ETD Abstract Container

Abstract Header

Activisee: Discrete and Continuous Activity Detection for Human Users Through Wearable Sensor-Augmented Glasses

Raychoudhury, Mrittika
http://orcid.org/0000-0001-8733-8925
http://rave.ohiolink.edu/etdc/view?acc_num=miami167239675075575

Abstract Details

2023, Master of Computer Science, Miami University, Computer Science and Software Engineering.
Presbyopia is the visual inability to focus on nearby objects and is common among older adults. Solutions include using separate glasses for different activities, like near-vision glasses for sedentary activities and distant-vision glasses for ambulatory activities. Forgetting to change eyeglasses when switching from sedentary to ambulatory activities, may lead to falls due to impaired depth perception and contrast sensitivity. Falls pose a serious risk to the well-being, confidence, and mortality of older people. In this research, we develop a novel machine learning (ML) and deep learning (DL) solution called Activisee, which uses sensor-augmented glasses to automatically detect user activities distinctly and to identify activity transitions. Although wireless sensor systems of other forms have long been used for fall detection and prevention through human activity recognition, using sensor-augmented glasses for activity transition detection has never previously been studied. Results using data collected from 28 human users performing a set of five distinct activities show that Activisee can achieve an accuracy of 89% using traditional ML algorithms and 95% using deep learning methods for activity detection. Also, Activisee when applied to continuous activity detection and transition recognition on real data can achieve a significantly high accuracy of up to 91%.
James Kiper (Advisor)
Jennifer Kinney (Committee Member)
Alan Ferrenberg (Committee Member)
Daniela Inclezan (Committee Member)
131 p.
Computer Science; Experiments; Technology
activity detection; smart-glasses; sensors; machine-learning; deep-learning;

Recommended Citations

Citations

  • Raychoudhury, M. (2023). Activisee: Discrete and Continuous Activity Detection for Human Users Through Wearable Sensor-Augmented Glasses [Master's thesis, Miami University]. OhioLINK Electronic Theses and Dissertations Center. http://rave.ohiolink.edu/etdc/view?acc_num=miami167239675075575

    APA Style (7th edition)

  • Raychoudhury, Mrittika. Activisee: Discrete and Continuous Activity Detection for Human Users Through Wearable Sensor-Augmented Glasses. 2023. Miami University, Master's thesis. OhioLINK Electronic Theses and Dissertations Center, http://rave.ohiolink.edu/etdc/view?acc_num=miami167239675075575.

    MLA Style (8th edition)

  • Raychoudhury, Mrittika. "Activisee: Discrete and Continuous Activity Detection for Human Users Through Wearable Sensor-Augmented Glasses." Master's thesis, Miami University, 2023. http://rave.ohiolink.edu/etdc/view?acc_num=miami167239675075575

    Chicago Manual of Style (17th edition)

miami167239675075575
25
© 2022, all rights reserved.
This open access ETD is published by Miami University and OhioLINK.