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

Classification of Systems-I01:26

Classification of Systems-I

700
Linearity is a system property characterized by a direct input-output relationship, combining homogeneity and additivity.
Homogeneity dictates that if an input x(t) is multiplied by a constant c, the output y(t) is multiplied by the same constant. Mathematically, this is expressed as:
700
Classification of Systems-II01:31

Classification of Systems-II

580
Continuous-time systems have continuous input and output signals, with time measured continuously. These systems are generally defined by differential or algebraic equations. For instance, in an RC circuit, the relationship between input and output voltage is expressed through a differential equation derived from Ohm's law and the capacitor relation,
580
Classification of Signals01:30

Classification of Signals

1.6K
In signal processing, signals are classified based on various characteristics: continuous-time versus discrete-time, periodic versus aperiodic, analog versus digital, and causal versus noncausal. Each category highlights distinct properties crucial for understanding and manipulating signals.
A continuous-time signal holds a value at every instant in time, representing information seamlessly. In contrast, a discrete-time signal holds values only at specific moments, often denoted as x(n), where...
1.6K

You might also read

Related Articles

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

Sort by
Same author

HRL-Det: Hierarchical Reinforcement Learning for Sequential Object Detection in Aerial Imagery.

Sensors (Basel, Switzerland)·2026
Same author

Programmable energy transfer in multivariate porphyrin-based 3D covalent organic frameworks.

Chemical communications (Cambridge, England)·2026
Same author

A prognostic model incorporating MCP-1 and IL-8 for early risk stratification in acute ischemic stroke patients receiving intravenous thrombolysis: a retrospective development with temporal validation.

Frontiers in neurology·2026
Same author

Association of food insecurity with metabolic dysfunction-associated steatotic liver disease by sex and race/ethnicity.

European journal of nutrition·2026
Same author

Integrated analysis on immunoreactivity and immunogenetics of Hantaan virus glycoprotein epitopes.

Microbial pathogenesis·2026
Same author

Deciphering small sequence differences in T cell receptor-antigen pairing.

Nature communications·2026

Related Experiment Video

Updated: Apr 15, 2026

Applying Hyperspectral Reflectance Imaging to Investigate the Palettes and the Techniques of Painters
07:05

Applying Hyperspectral Reflectance Imaging to Investigate the Palettes and the Techniques of Painters

Published on: June 18, 2021

2.9K

When Lie Groups Meet Hyperspectral Images: Equivariant Manifold Network for Few-Shot HSI Classification.

Haolong Ban1, Junchao Feng1, Zejin Liu1

  • 1College of Information Engineering, Dalian Ocean University, Dalian 116023, China.

Sensors (Basel, Switzerland)
|April 14, 2026
PubMed
Summary

EMNet, a novel Lie-group-based network, enhances few-shot hyperspectral image classification by encoding geometric invariance. This approach significantly improves accuracy and stability, even with limited data and complex disturbances.

Keywords:
Lie groupsSE(2) equivarianceaffine Lie groupfew-shot learninggeometric invariancehyperspectral image classificationmanifold modelingremote sensing

Related Experiment Videos

Last Updated: Apr 15, 2026

Applying Hyperspectral Reflectance Imaging to Investigate the Palettes and the Techniques of Painters
07:05

Applying Hyperspectral Reflectance Imaging to Investigate the Palettes and the Techniques of Painters

Published on: June 18, 2021

2.9K

Area of Science:

  • Remote Sensing
  • Computer Vision
  • Machine Learning

Background:

  • Hyperspectral imagery (HSI) provides rich spectral and spatial information but faces challenges in classification due to limited labeled data and geometric distortions (translation, rotation, scaling, shear).
  • Existing deep learning models often assume Euclidean structures and require extensive training data, proving ineffective in few-shot scenarios with complex geometric variations.

Purpose of the Study:

  • To develop EMNet, a Lie-group-based Equivariant Manifold Network designed to address the limitations of few-shot HSI classification.
  • To explicitly incorporate geometric invariance and enhance discriminative accuracy in HSI classification models, particularly under data scarcity and geometric disturbances.

Main Methods:

  • EMNet integrates an SE(2)-based Equivariance-Guided Module (EGM) to ensure equivariance to translations and rotations.
  • An affine Lie-group-based Characteristic Filtering Convolution (CFC) is employed to model scaling and shearing on the feature manifold while reducing redundant information.
  • The network is evaluated on multiple benchmark HSI datasets (WHU-Hi-HongHu, Houston2013, Indian Pines) and a large-scale city dataset (Xiongan New Area) under various few-shot protocols.

Main Results:

  • EMNet achieved state-of-the-art performance on benchmark datasets, with Overall Accuracies (OAs) reaching 95.77% (50 samples/class), 97.37% (50 samples/class), and 96.09% (5% labeled samples).
  • Significant improvements were observed compared to the DGPF-RENet baseline, including up to +3.34% OA, +6.01% AA, and +4.14% Kappa.
  • The model demonstrated enhanced stability and robustness, particularly on a large-scale dataset with extreme class imbalance (Xiongan New Area), boosting OA from 85.89% to 93.77% under a 1% labeled-sample protocol.

Conclusions:

  • EMNet effectively addresses the challenges of few-shot HSI classification by leveraging Lie-group theory for geometric invariance.
  • The proposed network offers a robust and stable solution for accurate HSI classification, even with scarce labels and complex geometric transformations.
  • EMNet shows strong potential for practical applications such as large-area HSI mapping and analysis.