Home Mathematics Chapter 12 Utilizing real-world data to develop a userindependent sensor-based human activity recognition system
Chapter
Licensed
Unlicensed Requires Authentication

Chapter 12 Utilizing real-world data to develop a userindependent sensor-based human activity recognition system

  • Enas E. Alkhoshi , Khaled M. Rasheed , Hamid R. Arabnia , Frederick W. Maier and Jennifer L. Gay
Become an author with De Gruyter Brill
Explore this Subject How to publish with us
Imaging Science
This chapter is in the book Imaging Science

Abstract

In human activity recognition (HAR), the selection of features that accurately reflect the data is vital for optimizing the performance of classification models. This study aimed to enhance the classification of metabolic equivalent of task (MET) levels of physical activities using body-worn sensor data by incorporating demographic and anthropometric features. We utilized a dataset from wearable accelerometers collected from 270 participants across various cities in Georgia, engaging in a range of physical activities. Our study revealed that integrating demographic and anthropometric features significantly improved model accuracy in classifying MET levels. We implemented a self-attention mechanism in a transformer model to analyze motion signals over time, capturing the nuanced relationships between signal levels within a time series. Additionally, we explored model personalization to address inter-subject variability, which notably outperformed the transformer model. By including only 30% of each participant’s data in the training set, we elevated the accuracy from 83.29% to 94.84%. This advancement underscores the challenge of achieving high performance in subject-independent systems when working with real-world data.

Abstract

In human activity recognition (HAR), the selection of features that accurately reflect the data is vital for optimizing the performance of classification models. This study aimed to enhance the classification of metabolic equivalent of task (MET) levels of physical activities using body-worn sensor data by incorporating demographic and anthropometric features. We utilized a dataset from wearable accelerometers collected from 270 participants across various cities in Georgia, engaging in a range of physical activities. Our study revealed that integrating demographic and anthropometric features significantly improved model accuracy in classifying MET levels. We implemented a self-attention mechanism in a transformer model to analyze motion signals over time, capturing the nuanced relationships between signal levels within a time series. Additionally, we explored model personalization to address inter-subject variability, which notably outperformed the transformer model. By including only 30% of each participant’s data in the training set, we elevated the accuracy from 83.29% to 94.84%. This advancement underscores the challenge of achieving high performance in subject-independent systems when working with real-world data.

Chapters in this book

  1. Frontmatter I
  2. Preface V
  3. Contents VII
  4. Section: Image processing
  5. Chapter 1 Magnetic resonance image re-parameterization on real data 1
  6. Chapter 2 Denoising and gradient fusion for effective edge detection for noisy color images 17
  7. Chapter 3 Understanding driver attention to objects for ADASs: what do drivers see? 39
  8. Chapter 4 Image clustering enhanced with refined image classification 59
  9. Chapter 5 AI-powered framework for objective scoring of product design innovation 89
  10. Section: Computer vision
  11. Chapter 6 Image inpainting using GAN transformerbased model 111
  12. Chapter 7 Enhanced image watermarking through cross-attention and noise-invariant domain learning 127
  13. Chapter 8 Online melt pool monitoring using a deep transformer image processing solution 153
  14. Chapter 9 Implementation of deep learning techniques on thermal image classification 173
  15. Chapter 10 Drishti: a generative AI-based application for gesture recognition and execution 203
  16. Section: Pattern recognition
  17. Chapter 11 Exploring muzzle biometrics: a deep learning framework for noninvasive cattle recognition 239
  18. Chapter 12 Utilizing real-world data to develop a userindependent sensor-based human activity recognition system 253
  19. Index 273
https://doi.org/10.1515/9783111436425-012

Chapters in this book

  1. Frontmatter I
  2. Preface V
  3. Contents VII
  4. Section: Image processing
  5. Chapter 1 Magnetic resonance image re-parameterization on real data 1
  6. Chapter 2 Denoising and gradient fusion for effective edge detection for noisy color images 17
  7. Chapter 3 Understanding driver attention to objects for ADASs: what do drivers see? 39
  8. Chapter 4 Image clustering enhanced with refined image classification 59
  9. Chapter 5 AI-powered framework for objective scoring of product design innovation 89
  10. Section: Computer vision
  11. Chapter 6 Image inpainting using GAN transformerbased model 111
  12. Chapter 7 Enhanced image watermarking through cross-attention and noise-invariant domain learning 127
  13. Chapter 8 Online melt pool monitoring using a deep transformer image processing solution 153
  14. Chapter 9 Implementation of deep learning techniques on thermal image classification 173
  15. Chapter 10 Drishti: a generative AI-based application for gesture recognition and execution 203
  16. Section: Pattern recognition
  17. Chapter 11 Exploring muzzle biometrics: a deep learning framework for noninvasive cattle recognition 239
  18. Chapter 12 Utilizing real-world data to develop a userindependent sensor-based human activity recognition system 253
  19. Index 273
Sign up now to receive a 20% welcome discount
Institutional Access
How does access work?
Have an idea on how to improve our website?
Please write us.
© 2025 De Gruyter Brill
Downloaded on 20.11.2025 from https://www.degruyterbrill.com/document/doi/10.1515/9783111436425-012/html
Scroll to top button