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

Successful Immunosuppressive Therapy for Aseptic Abscess Syndrome Presenting with a Fever of Unknown Origin.

Internal medicine (Tokyo, Japan)·2026
Same author

Individual Variability in Deep Learning-Based Joint Angle Estimation from a Single IMU: A Cross-Population Study.

Sensors (Basel, Switzerland)·2026
Same author

Influence of depopulation on pollutant loads generated in the Seto inland sea basin.

Scientific reports·2025
Same author

Comprehensive Molecular Studies in 88 Japanese Patients With Congenital Hypogonadotropic Hypogonadism.

The Journal of clinical endocrinology and metabolism·2025
Same author

A Deep Learning Approach for Biped Robot Locomotion Interface Using a Single Inertial Sensor.

Sensors (Basel, Switzerland)·2023
Same author

Comparison of physician- and self-assessed pubertal onset in Japanese children.

Frontiers in pediatrics·2023

Related Experiment Video

Updated: Nov 7, 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

4.7K

New Sensor Data Structuring for Deeper Feature Extraction in Human Activity Recognition.

Tsige Tadesse Alemayoh1, Jae Hoon Lee1, Shingo Okamoto1

  • 1Department of Mechanical Engineering, Graduate School of Science and Engineering, Ehime University, Matsuyama 790-8577, Japan.

Sensors (Basel, Switzerland)
|April 30, 2021
PubMed
Summary

This study introduces a novel method for structuring time series data to improve human activity recognition using smartphones. A deep learning approach achieved high accuracy, demonstrating practical real-time applications.

Keywords:
convolutional neural networkdeep learninghuman activity recognitioninertial measurement unit sensorsinput adaptation

More Related Videos

Design and Evaluation of Smart Glasses for Food Intake and Physical Activity Classification
07:47

Design and Evaluation of Smart Glasses for Food Intake and Physical Activity Classification

Published on: February 14, 2018

11.5K
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

9.1K

Related Experiment Videos

Last Updated: Nov 7, 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

4.7K
Design and Evaluation of Smart Glasses for Food Intake and Physical Activity Classification
07:47

Design and Evaluation of Smart Glasses for Food Intake and Physical Activity Classification

Published on: February 14, 2018

11.5K
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

9.1K

Area of Science:

  • Computer Science
  • Artificial Intelligence
  • Human-Computer Interaction

Background:

  • Effective human-assistive technologies require computing devices to accurately recognize human movements in real-time.
  • Ubiquitous sensors and intelligent methods are crucial for developing reliable activity recognition systems.

Purpose of the Study:

  • To develop a novel data structuring method for deeper feature extraction in human activity recognition.
  • To enhance the performance of real-time activity recognition for smart devices and human-robot collaboration.

Main Methods:

  • Collected activity data from eight activities using a smartphone and a custom iOS application.
  • Structured data into single and double channels for deep temporal and spatial feature extraction.
  • Utilized Fourier and wavelet domains alongside the time domain for data representation.
  • Employed various neural network models, focusing on a convolutional neural network with double-channeled time-domain input.

Main Results:

  • The convolutional neural network with double-channeled time-domain input demonstrated superior performance in activity classification.
  • The proposed method achieved better performance when evaluated on public datasets.
  • Real-time testing on computers and smartphones showed promising results for the trained model's practicability.

Conclusions:

  • The developed data structuring and deep learning method significantly improves human activity recognition accuracy.
  • This approach offers a practical and efficient solution for real-time human movement awareness in smart devices and assistive technologies.