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: May 5, 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

1.3K

CSFCNet: Cascaded Spatial-Frequency Convolutional Network for Hyperspectral Image Classification.

Feng Jiang1,2, Xin Liu1, Mingxuan Li1

  • 1Changchun Institute of Optics, Fine Mechanics and Physics, Chinese Academy of Sciences, Changchun 130033, China.

Sensors (Basel, Switzerland)
|May 4, 2026
PubMed
Summary

Related Concept Videos

Cascaded Op Amps01:16

Cascaded Op Amps

1.3K
Operational amplifiers (op-amps) are versatile electronic components that can be interconnected in a cascade - one after another in a linear sequence. This cascading is possible due to their infinite input resistance and zero output resistance, allowing them to maintain their input-output relationships even when connected in series.
In a cascaded system, each op-amp is referred to as a stage. The output of one stage drives the input of the subsequent stage. As the input signal passes through...
1.3K

You might also read

Related Articles

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

Sort by
Same author

Ratiometric fluorometric and visual determination of cyanide based on the use of carbon dots and gold nanoclusters.

Mikrochimica acta·2019
Same author

Preparation and characterization of chitosan from crab shell (Portunus trituberculatus) by NaOH/urea solution freeze-thaw pretreatment procedure.

International journal of biological macromolecules·2019
Same author

Construction and application of an HSP70 promoter-inducible genome editing system in transgenic silkworm to induce resistance to Nosema bombycis.

Applied microbiology and biotechnology·2019
Same author

SNHG14 stimulates cell autophagy to facilitate cisplatin resistance of colorectal cancer by regulating miR-186/ATG14 axis.

Biomedicine & pharmacotherapy = Biomedecine & pharmacotherapie·2019
Same author

Relationship between left ventricular isovolumic relaxation flow patterns and mitral inflow patterns studied by using vector flow mapping.

Scientific reports·2019
Same author

A Microfluidic Device Integrating Impedance Flow Cytometry and Electric Impedance Spectroscopy for High-Efficiency Single-Cell Electrical Property Measurement.

Analytical chemistry·2019

The Cascaded Spatial-Frequency Convolutional Network (CSFCNet) enhances hyperspectral image classification by integrating spatial and frequency domain information. This novel approach improves feature extraction and addresses class imbalance for better accuracy.

Area of Science:

  • Remote Sensing
  • Computer Vision
  • Signal Processing

Background:

  • Convolutional Neural Networks (CNNs) excel at feature extraction in hyperspectral image classification but struggle with limited receptive fields, hindering multi-scale and global context capture.
  • Class imbalance in hyperspectral datasets often biases models, reducing overall classification accuracy by disproportionately weighting certain spectral bands.

Purpose of the Study:

  • To propose the Cascaded Spatial-Frequency Convolutional Network (CSFCNet) for improved hyperspectral image classification.
  • To address the limitations of CNNs in capturing multi-scale structural and global contextual information.
  • To mitigate the impact of class imbalance in hyperspectral image datasets.

Main Methods:

  • Developed a Dual Spatial Fourier Convolution (DSF-Conv) module to jointly model spatial and frequency domains, projecting feature maps into parallel representations.
Keywords:
Fourier transformhyperspectral image classificationlocal attention mechanismspatial convolution

Related Experiment Videos

Last Updated: May 5, 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

1.3K
  • Employed multi-scale convolutions in the spatial pathway for hierarchical structure extraction and frequency-domain convolutions in the Fourier pathway for global context aggregation.
  • Integrated a group-cascaded structure with residual connections to alleviate class imbalance and introduced a Lightweight Local Attention module for enhanced feature discrimination.

    • Main Results:

    • CSFCNet achieved competitive accuracies on three benchmark hyperspectral image datasets.
    • The proposed method effectively integrates spatial and frequency domain information, overcoming CNN limitations.
    • Ablation studies confirmed the significant contribution of the core components, including DSF-Conv and the attention module.

    Conclusions:

    • CSFCNet demonstrates superior performance in hyperspectral image classification by effectively leveraging multi-domain information.
    • The network architecture successfully addresses challenges related to feature extraction and class imbalance.
    • The findings highlight the potential of integrating spatial and frequency domain analysis for advanced hyperspectral imaging applications.