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 Experiment Videos

Intelligent Virtual Sensor Generation Using KL-Divergence- Based Fusion and Deep Generative Learning for Smart

Murad Ali Khan1, Qazi Waqas Khan1, Muhammad Faizan1

  • 1Department of Computer Engineering, Jeju National University, Jeju 63243, Republic of Korea.

Sensors (Basel, Switzerland)
|July 15, 2026
PubMed
Summary

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

Ruminants Epidemiology, Spatial Clustering and Multivariate Risk Modelling of Cutaneous Leishmaniasis in District Bajaur, Pakistan.

Veterinary medicine and science·2026
Same author

Integrated diagnostic and host biomarker assessment of plasmodium infections in malaria-endemic Peshawar, Pakistan.

Acta tropica·2026
Same author

Prevalence, Species Diversity, and Risk Factors of Tick Infestation in Cattle From District Peshawar, Pakistan.

Veterinary medicine and science·2025
Same author

Fog-assisted de-duplicated data exchange in distributed edge computing networks.

Scientific reports·2024
Same author

A Comparative Study on Imputation Techniques: Introducing a Transformer Model for Robust and Efficient Handling of Missing EEG Amplitude Data.

Bioengineering (Basel, Switzerland)·2024
Same author

An intelligent diabetes classification and perception framework based on ensemble and deep learning method.

PeerJ. Computer science·2024

This study introduces a virtual sensor framework to improve environmental monitoring data quality. The intelligent system generates reliable data, overcoming sensor faults and sparse deployment for better environmental insights.

Area of Science:

  • Environmental Science
  • Data Science
  • Sensor Technology

Background:

  • Sensor-based environmental monitoring systems frequently suffer from data quality issues like missing values, noise, and unreliability due to sensor faults, sparse deployment, calibration drift, and communication interruptions.
  • These data deficiencies hinder accurate environmental analysis and decision-making, necessitating robust solutions for data imputation and enhancement.

Purpose of the Study:

  • To propose an intelligent virtual sensor generation framework designed to address data quality challenges in environmental monitoring.
  • To develop a comprehensive system that integrates preprocessing, statistical modeling, data fusion, deep generative augmentation, and temporal prediction for virtual sensor data.

Main Methods:

  • Physical-constraint-based preprocessing including threshold filtering, validity checks, and outlier detection (Isolation Forest).
Keywords:
BiGRUBiLSTMKL divergenceconditional tabular GANdeep generative learningenvironmental monitoringintelligent sensingmachine learning for sensor data analysissensor data fusionsmart sensorsvariational autoencodervirtual sensors

Related Experiment Videos

  • Statistical virtual sensor modeling (IDW, KDE, Ridge Regression, Copula) and KL-divergence-based adaptive fusion of generated data.
  • Deep generative augmentation using Variational Autoencoders (VAE) and Conditional Tabular Generative Adversarial Networks (CTGAN), followed by temporal prediction (BiLSTM, BiGRU).

    • Main Results:

    • The proposed framework successfully generates physically valid and distributionally consistent virtual sensor data.
    • Fusion-based methods demonstrated superior performance over standalone approaches.
    • VAE-based augmentation achieved better statistical fidelity and lower prediction errors compared to CTGAN.
    • The framework showed significant performance and transferability on a public NOAA dataset, outperforming other generative baselines.

    Conclusions:

    • The developed intelligent virtual sensor framework provides a reliable and scalable solution for enhancing data quality in sensor-sparse and fault-prone environmental monitoring systems.
    • The integration of preprocessing, fusion, and deep generative methods offers a robust approach to virtual sensing, improving data completeness and accuracy.
    • The framework's computational feasibility supports its practical application in real-world environmental monitoring scenarios.