Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

A Novel Traffic Scheduling Algorithm for Multi-CQF Using Mixed Integer Programming and Variable Neighborhood Search Genetic Algorithm in Time-Sensitive Networking.

Sensors (Basel, Switzerland)·2025
Same author

Static-Dynamic Temporal Networks for Parkinson's Disease Detection and Severity Prediction.

IEEE transactions on neural systems and rehabilitation engineering : a publication of the IEEE Engineering in Medicine and Biology Society·2023
Same author

Human Behavior Recognition Model Based on Feature and Classifier Selection.

Sensors (Basel, Switzerland)·2021
Same author

[Clinical significance of changes in T wave and ST segment amplitudes on electrocardiogram from supine to standing position among children with unexplained chest tightness or pain in resting stage].

Zhongguo dang dai er ke za zhi = Chinese journal of contemporary pediatrics·2013
Same author

Biomimetic synthesis of equisetin and (+)-fusarisetin A.

Chemistry (Weinheim an der Bergstrasse, Germany)·2013
Same author

ERG Protein Expression Is of Limited Prognostic Value in Men with Localized Prostate Cancer.

ISRN urology·2013

Related Experiment Video

Updated: Aug 25, 2025

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention
06:37

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention

Published on: December 15, 2023

4.0K

A Novel Segmentation Scheme with Multi-Probability Threshold for Human Activity Recognition Using Wearable Sensors.

Bangwen Zhou1, Cheng Wang1, Zhan Huan2

  • 1School of Computer Science and Artificial Intelligence, Aliyun School of Big Data, School of Software, Changzhou University, Changzhou 213000, China.

Sensors (Basel, Switzerland)
|October 14, 2022
PubMed
Summary

This study introduces a new machine learning method for human activity recognition (HAR) using wearable sensors. The novel approach effectively filters out unwanted activities, significantly improving recognition accuracy.

Keywords:
human activity recognitionmachine learningmulti-label weighted probabilityslope-area methodthreshold segmentation

More Related Videos

Design and Analysis for Fall Detection System Simplification
08:05

Design and Analysis for Fall Detection System Simplification

Published on: April 6, 2020

10.8K
Evaluation of a Smartphone-based Human Activity Recognition System in a Daily Living Environment
06:49

Evaluation of a Smartphone-based Human Activity Recognition System in a Daily Living Environment

Published on: December 11, 2015

9.0K

Related Experiment Videos

Last Updated: Aug 25, 2025

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention
06:37

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention

Published on: December 15, 2023

4.0K
Design and Analysis for Fall Detection System Simplification
08:05

Design and Analysis for Fall Detection System Simplification

Published on: April 6, 2020

10.8K
Evaluation of a Smartphone-based Human Activity Recognition System in a Daily Living Environment
06:49

Evaluation of a Smartphone-based Human Activity Recognition System in a Daily Living Environment

Published on: December 11, 2015

9.0K

Area of Science:

  • Computer Science
  • Biomedical Engineering
  • Machine Learning

Background:

  • Human Activity Recognition (HAR) using wearable sensors is crucial but challenged by manual labeling and complex time-series signals.
  • Existing HAR methods struggle with filtering transition and atypical activities within complete time-series data.
  • The presence of diverse activities, including static, dynamic, and transitional ones, complicates accurate HAR.

Purpose of the Study:

  • To propose a novel machine learning-based segmentation scheme for Human Activity Recognition (HAR).
  • To address the challenge of accurately filtering transition and atypical activities in time-series sensor data.
  • To reduce HAR errors caused by fixed windows and improve the rejection of unknown activities.

Main Methods:

  • A novel machine learning-based segmentation scheme employing a multi-probability threshold is introduced.
  • Threshold Segmentation (TS) and Slope-Area (SA) approaches are utilized, tailored to signal characteristics.
  • A Multi-Label Weighted Probability (MLWP) model is developed for activity probability estimation.

Main Results:

  • The proposed model significantly decreases HAR errors by overcoming fixed-window limitations.
  • Unknown activities are accurately rejected, minimizing their impact on recognition.
  • High performance was demonstrated on the UCI and PAMAP2 datasets, achieving average HAR accuracies of 97.71% and 95.93% respectively.

Conclusions:

  • The novel segmentation scheme and MLWP model offer a robust solution for HAR.
  • The method effectively handles complex time-series data, improving accuracy and reliability.
  • This approach advances HAR by accurately recognizing activities and filtering noise from sensor data.