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 Video

Updated: Jul 16, 2026

End-To-End Deep Neural Network for Salient Object Detection in Complex Environments
03:31

End-To-End Deep Neural Network for Salient Object Detection in Complex Environments

Published on: December 15, 2023

MDC-MobileNetV3: A Lightweight Multi-Scale Hierarchical Attention Network for Remote Sensing Scene Classification.

Haonan Liu1, Xiao Wang1, Jialong Sun1

  • 1School of Marine Technology and Geomatics, Jiangsu Ocean University, Lianyungang 222005, China.

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

Halogen-Electronegativity Tuning Induces Symmetry Breaking and Polarity Activation in Chiral Zinc Halide Hybrids.

Inorganic chemistry·2026
Same author

Dynamic changes in In vitro rumen fermentation, nutrient degradation, and microbial communities of fermentation inoculant-treated licorice stem and leaf silage under different dry matter contents.

PloS one·2026
Same author

Comparative Effectiveness and Safety of Interventions for Atrial Fibrillation in Heart Failure: A Network Meta-Analysis of Randomized Trials.

Cardiovascular therapeutics·2026
Same author

Fecal Microbiota Transplantation Improves Biota and Hepatic Metabolism, Promoting Growth in SD Rats Under Hypobaric Hypoxia Exposure.

Microorganisms·2026
Same author

EDM-Net: A Multi-Scale Network for Object Detection in Remote Sensing Images.

Sensors (Basel, Switzerland)·2026
Same author

TkNACs Heterodimerization and Methyl Jasmonate Signaling Synergistically Mediate Root Development in <i>Taraxacum kok-saghyz</i>.

Plants (Basel, Switzerland)·2026

A new lightweight framework, MDC-MobileNetV3, improves remote sensing scene classification by extracting multi-scale features and adaptively recalibrating them. This approach enhances accuracy while maintaining a compact model size.

Area of Science:

  • Computer Science
  • Remote Sensing
  • Artificial Intelligence

Background:

  • Remote sensing scene classification faces challenges from object scale variations, background interference, and inter-class similarity.
  • Existing methods often struggle to balance accuracy and model efficiency.

Purpose of the Study:

  • To propose a lightweight and effective framework for remote sensing scene classification.
  • To address the limitations of existing methods in handling complex remote sensing data.

Main Methods:

  • Developed MDC-MobileNetV3, a lightweight classification framework using the MobileNetV3-Large backbone.
  • Integrated a Multi-Scale Feature Extraction (MSFE) module for diverse spatial information capture.
  • Incorporated a Dynamic Feature Weighted Fusion (DFWF) mechanism for adaptive feature recalibration.
Keywords:
DFWFMSFEMobileNetV3remote sensing scene classification

Related Experiment Videos

Last Updated: Jul 16, 2026

End-To-End Deep Neural Network for Salient Object Detection in Complex Environments
03:31

End-To-End Deep Neural Network for Salient Object Detection in Complex Environments

Published on: December 15, 2023

  • Employed a hierarchical CBAM attention strategy to enhance discriminative region representation.

    • Main Results:

    • Achieved high classification accuracies across multiple benchmark datasets (e.g., 99.52% on UC Merced, 96.48% on NWPU-Resisc45).
    • Maintained a lightweight architecture with approximately 4.35 million parameters.
    • Grad-CAM visualizations confirmed the model's focus on semantically meaningful regions and suppression of irrelevant background information.

    Conclusions:

    • The proposed MDC-MobileNetV3 framework offers a favorable trade-off between classification accuracy and model lightweight design.
    • The model demonstrates effectiveness in remote sensing scene understanding, providing improved interpretability.