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

Adolescents' cross-cultural aesthetic preferences are shaped by cultural background, art exposure, and developmental psychological traits.

Frontiers in psychology·2026
Same author

State-Aware Resource Allocation for V2X Communications.

Sensors (Basel, Switzerland)·2026
Same author

Channel and Power Allocation for Multi-Cell NOMA Using Multi-Agent Deep Reinforcement Learning and Unsupervised Learning.

Sensors (Basel, Switzerland)·2025
Same author

Cine clips increase the detection of thyroid pyramidal lobe in routine thyroid sonogram.

Ultrasound (Leeds, England)·2024
Same author

Joint Deep Reinforcement Learning and Unsupervised Learning for Channel Selection and Power Control in D2D Networks.

Entropy (Basel, Switzerland)·2022

Related Experiment Video

Updated: Jun 7, 2025

Author Spotlight: UAV Remote Sensing for Efficient Invasive Plant Biomass Estimation
08:47

Author Spotlight: UAV Remote Sensing for Efficient Invasive Plant Biomass Estimation

Published on: February 9, 2024

1.3K

Bidirectional Mamba with Dual-Branch Feature Extraction for Hyperspectral Image Classification.

Ming Sun1,2, Jie Zhang1, Xiaoou He1

  • 1College of Computer and Control Engineering, Qiqihar University, Qiqihar 161006, China.

Sensors (Basel, Switzerland)
|November 9, 2024
PubMed
Summary

This study introduces DBMamba, a novel CNN-Mamba network for hyperspectral image (HSI) classification. DBMamba effectively utilizes spectral features and reduces computational costs, outperforming existing methods.

Keywords:
CNNHSIMambadual-branch feature extraction

More Related Videos

Excitation-Scanning Hyperspectral Imaging Microscopy to Efficiently Discriminate Fluorescence Signals
00:07

Excitation-Scanning Hyperspectral Imaging Microscopy to Efficiently Discriminate Fluorescence Signals

Published on: August 22, 2019

8.0K
Dual-mode Imaging of Cutaneous Tissue Oxygenation and Vascular Function
11:35

Dual-mode Imaging of Cutaneous Tissue Oxygenation and Vascular Function

Published on: December 8, 2010

16.5K

Related Experiment Videos

Last Updated: Jun 7, 2025

Author Spotlight: UAV Remote Sensing for Efficient Invasive Plant Biomass Estimation
08:47

Author Spotlight: UAV Remote Sensing for Efficient Invasive Plant Biomass Estimation

Published on: February 9, 2024

1.3K
Excitation-Scanning Hyperspectral Imaging Microscopy to Efficiently Discriminate Fluorescence Signals
00:07

Excitation-Scanning Hyperspectral Imaging Microscopy to Efficiently Discriminate Fluorescence Signals

Published on: August 22, 2019

8.0K
Dual-mode Imaging of Cutaneous Tissue Oxygenation and Vascular Function
11:35

Dual-mode Imaging of Cutaneous Tissue Oxygenation and Vascular Function

Published on: December 8, 2010

16.5K

Area of Science:

  • Remote Sensing
  • Computer Vision
  • Machine Learning

Background:

  • Hyperspectral image (HSI) classification is crucial for remote sensing analysis.
  • Convolutional Neural Networks (CNNs) have advanced HSI classification but struggle with spectral feature utilization and computational costs.
  • Existing methods often fail to fully capture the sequential spectral properties of HSI data.

Purpose of the Study:

  • To propose a novel network, DBMamba, that addresses the limitations of CNNs in HSI classification.
  • To enhance the utilization of spectral features and reduce computational complexity in HSI classification.
  • To improve the overall accuracy and efficiency of hyperspectral image classification.

Main Methods:

  • A CNN-Mamba architecture (DBMamba) is proposed, integrating CNNs with a bidirectional Mamba.
  • Principal Component Analysis (PCA) is employed for initial feature extraction.
  • A dual-branch CNN (3D-CNN and 2D-CNN) extracts shallow spectral-spatial features, followed by a bidirectional Mamba for global contextual information and spectral feature enhancement.

Main Results:

  • DBMamba demonstrates superior classification performance on benchmark datasets (Indian Pines, Salinas, Pavia University).
  • The proposed method achieved accuracy improvements of 1.04%, 0.15%, and 0.09% over the SSFTT method on the respective datasets.
  • The network effectively captures global contextual information and enhances spectral feature extraction with linear computational cost.

Conclusions:

  • DBMamba offers a significant advancement in hyperspectral image classification by effectively leveraging spectral-spatial features and Mamba's sequential processing capabilities.
  • The proposed architecture provides a more efficient and accurate approach compared to existing state-of-the-art methods.
  • This research contributes valuable insights for future developments in deep learning-based HSI classification.