Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Process-aware Human Activity Recognition

Jiawei Zheng, Petros Papapanagiotou, Jacques D. Fleuriot, Jane Hillston

TL;DR

This work proposes a novel approach that incorporates process information from context to enhance the HAR performance, and aligns probabilistic events generated by machine learning models with process models derived from contextual information.

Abstract

Humans naturally follow distinct patterns when conducting their daily activities, which are driven by established practices and processes, such as production workflows, social norms and daily routines. Human activity recognition (HAR) algorithms usually use neural networks or machine learning techniques to analyse inherent relationships within the data. However, these approaches often overlook the contextual information in which the data are generated, potentially limiting their effectiveness. We propose a novel approach that incorporates process information from context to enhance the HAR performance. Specifically, we align probabilistic events generated by machine learning models with process models derived from contextual information. This alignment adaptively weighs these two sources of information to optimise HAR accuracy. Our experiments demonstrate that our approach achieves better accuracy and Macro F1-score compared to baseline models.

Process-aware Human Activity Recognition

TL;DR

This work proposes a novel approach that incorporates process information from context to enhance the HAR performance, and aligns probabilistic events generated by machine learning models with process models derived from contextual information.

Abstract

Humans naturally follow distinct patterns when conducting their daily activities, which are driven by established practices and processes, such as production workflows, social norms and daily routines. Human activity recognition (HAR) algorithms usually use neural networks or machine learning techniques to analyse inherent relationships within the data. However, these approaches often overlook the contextual information in which the data are generated, potentially limiting their effectiveness. We propose a novel approach that incorporates process information from context to enhance the HAR performance. Specifically, we align probabilistic events generated by machine learning models with process models derived from contextual information. This alignment adaptively weighs these two sources of information to optimise HAR accuracy. Our experiments demonstrate that our approach achieves better accuracy and Macro F1-score compared to baseline models.

Paper Structure

This paper contains 18 sections, 3 equations, 4 figures, 7 tables.

Table of Contents

  1. Introduction
  2. Motivating example
  3. Preliminary
  4. Method
  5. Extracting probabilistic traces
  6. Process discovery
  7. Alignment
  8. Experimental study
  9. Dataset description
  10. Experimental design
  11. Process discovery
  12. Results and discussion
  13. Related work
  14. Conclusion
  15. Appendix
  16. ...and 3 more sections

Figures (4)

  • Figure 1: Example of a process model.
  • Figure 2: The framework of process-driven human activity recognition.
  • Figure 3: Discovered process models. The number in the node label indicates the frequency of the corresponding activity. The number in the arc indicates the frequency of the corresponding path.
  • Figure 4: Accuracy of activity recognition across varying confidence thresholds $\epsilon$ in the validation set.