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

Force Classification01:22

Force Classification

2.7K
Forces play a crucial role in the study of physics and engineering. They are essential in describing the motion, behavior, and equilibrium of objects in the physical world. Forces can be classified based on their origin, type, and direction of action.
Contact and non-contact forces are two of the most widely used categories of forces. As the name suggests, contact forces require physical contact between two objects to act upon each other. Examples of contact forces include frictional,...
2.7K

You might also read

Related Articles

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

Sort by
Same author

Stock Forecasting Based on Informational Complexity Representation: A Framework of Wavelet Entropy, Multiscale Entropy, and Dual-Branch Network.

Entropy (Basel, Switzerland)·2026
Same author

Fusing Geometric and Semantic Features via Cosine Similarity Cross-Attention for Remote Sensing Scene Classification.

Sensors (Basel, Switzerland)·2026
Same author

A Multi-Scale Object Detection Network with Integrated Spatial-Channel Collaborative Attention for Remote Sensing Images.

Sensors (Basel, Switzerland)·2026
See all related articles

Related Experiment Video

Updated: Apr 8, 2026

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
03:31

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Published on: December 15, 2023

1.2K

Towards Lightweight and Multi-Scale Scene Classification: A Lie Group-Guided Deep Learning Network with Collaborative

Xuefei Xu1, Chengjun Xu2,3

  • 1School of Information Engineering, Shanghai Dianji University, Shanghai 201306, China.

Journal of Imaging
|March 27, 2026
PubMed
Summary

This study introduces the Lie Group lightweight multi-scale network (LGLMNet) for remote sensing scene classification. LGLMNet enhances accuracy by integrating shallow and deep features, offering a lightweight yet effective solution.

Keywords:
Lie Group machine learningattention mechanismfeature fusionremote sensing scene classification

More Related Videos

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

5.6K

Related Experiment Videos

Last Updated: Apr 8, 2026

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
03:31

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Published on: December 15, 2023

1.2K
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

5.6K

Area of Science:

  • Earth observation
  • Computer vision
  • Machine learning

Background:

  • Current deep learning methods for remote sensing scene classification (RSSC) prioritize high-level semantics, neglecting crucial shallow details like edges and textures.
  • Conventional Convolutional Neural Networks (CNNs) have fixed receptive fields, and Transformers are computationally expensive, limiting their application in RSSC.

Purpose of the Study:

  • To develop a lightweight and accurate network for remote sensing scene classification (RSSC).
  • To address the limitations of existing methods by integrating complementary shallow and deep features.
  • To improve the efficiency and performance of Earth observation data analysis.

Main Methods:

  • Proposed the Lie Group lightweight multi-scale network (LGLMNet), a dual-branch architecture combining Lie Group machine learning (LGML) for shallow features and a deep learning branch for semantics.
  • Incorporated a parallel depthwise separable convolution block (PDSCB) for multi-scale perception and a spatial-channel collaborative attention mechanism (SCCA) for global-local modeling within the deep branch.
  • Utilized a cross-layer feature fusion block (CLFFB) to effectively merge features from both branches.

Main Results:

  • LGLMNet achieved accuracy improvements of 2.14% on UCM-21, 2.32% on AID, and 1.12% on NWPU-45 datasets compared to state-of-the-art methods.
  • The proposed network maintains a lightweight structure with only 2.6 million parameters, demonstrating high efficiency.
  • The integration of Lie Group covariance features effectively captures both shallow and deep information for improved classification.

Conclusions:

  • LGLMNet offers a superior approach to remote sensing scene classification by effectively combining shallow and deep features.
  • The lightweight design and improved accuracy make LGLMNet a promising tool for practical Earth observation applications.
  • The study highlights the potential of Lie Group machine learning in enhancing deep learning models for complex visual recognition tasks.