Jove
Visualize
Contact Us

Related Experiment Video

Updated: Mar 15, 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

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

Xuefei Xu1, Chengjun Xu2,3

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

Sensors (Basel, Switzerland)
|March 14, 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

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

Entropy (Basel, Switzerland)·2026
Same author

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

Journal of imaging·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
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

This study introduces CBCAM-LGM, an efficient framework for high-resolution remote sensing image scene classification. It achieves state-of-the-art accuracy with significantly reduced complexity by fusing multi-level features adaptively.

Area of Science:

  • Earth Observation
  • Computer Vision
  • Artificial Intelligence

Background:

  • High-resolution remote sensing image scene classification (HRRSI-SC) is vital for Earth surface information but faces challenges like background interference and feature ambiguity.
  • Convolutional Neural Networks (CNNs) struggle with long-range dependencies, while Vision Transformers (ViTs) are computationally intensive and may lack local feature modeling.
  • Existing methods often fail to balance global and local feature extraction efficiently, leading to suboptimal performance in HRRSI-SC.

Purpose of the Study:

  • To propose an efficient and accurate cross-level feature complementary classification framework, CBCAM-LGM, for HRRSI-SC.
  • To address the limitations of CNNs and ViTs by enabling adaptive fusion of multi-granularity features.
  • To reduce computational complexity while maintaining high classification accuracy.
Keywords:
Lie Groupcross-level attentiondual-branch networkmulti-scale fusionremote sensing scene classification

Related Experiment Videos

Last Updated: Mar 15, 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

Main Methods:

  • A novel cross-level bidirectional complementary attention module (CBCAM) for adaptive feature fusion using cross-query attention.
  • Distillation of multi-granularity features via global average pooling to reduce redundancy.
  • Parallel dilated convolutions with parameter sharing to capture multi-scale contextual information efficiently.

Main Results:

  • Achieved state-of-the-art classification accuracy of 97.81% on the AID dataset, surpassing ViT-B-16 by 1.63%.
  • Reduced computational complexity to 1.21 GMACs with only 11.237 million parameters (an 87% reduction).
  • Demonstrated superior performance in capturing multi-scale contextual information and handling feature ambiguities.

Conclusions:

  • The proposed CBCAM-LGM framework offers an efficient and highly accurate solution for HRRSI-SC.
  • The cross-level feature complementary approach effectively balances global and local feature extraction.
  • The model presents a significant advancement in terms of accuracy, parameter efficiency, and computational cost for remote sensing image analysis.