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 Videos

A multiscale expectation-maximization semisupervised classifier suitable for badly posed image classification.

Andrea Baraldi1, Lorenzo Bruzzone, Palma Blonda

  • 1Istituto di Studi su Sistemi Intelligenti per l'Automazione, Consiglio Nazionale delle Ricerche, 70126 Bari, Italy. baraldi@ba.issia.cnr.it

IEEE Transactions on Image Processing : a Publication of the IEEE Signal Processing Society
|August 12, 2006
PubMed
Summary
This summary is machine-generated.

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

Radar-based observation of a lava tube on Venus.

Nature communications·2026
Same author

REST: Holistic Learning for End-to-End Semantic Segmentation of Whole-Scene Remote Sensing Imagery.

IEEE transactions on pattern analysis and machine intelligence·2025
Same author

CrossDiff: Exploring Self-SupervisedRepresentation of Pansharpening via Cross-Predictive Diffusion Model.

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2024
Same author

Characterization of the Surfaces and Near-Surface Atmospheres of Ganymede, Europa and Callisto by JUICE.

Space science reviews·2024
Same author

Raman Spectroscopy as an Effective Tool for Assessment of Structural Quality and Polymorphism of Gallium Oxide (Ga<sub>2</sub>O<sub>3</sub>) Thin Films.

Applied spectroscopy·2024
Same author

Radar for Europa Assessment and Sounding: Ocean to Near-Surface (REASON).

Space science reviews·2024
Same journal

SinColor: Uncertainty-Guided Single-Step Diffusion for Image Colorization.

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2026
Same journal

Through the Looking Glass: A Dual Perspective on Weakly-Supervised Few-Shot Segmentation.

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2026
Same journal

Mask-guided Asymmetric Contrastive and Semantic Alignment for Unsupervised Person Re-Identification.

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2026
Same journal

Hyperbolic Cycle Alignment for Infrared-Visible Image Fusion.

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2026
Same journal

Learning Gaze Synthesizer via 3D-eye Controlled Diffusion and Cross-domain Feature Alignment.

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2026
Same journal

Underlying Semantic Diffusion for Effective and Efficient In-Context Learning.

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2026
See all related articles

A new multiscale semisupervised expectation maximization (MSEM) classifier addresses poorly defined image classification tasks with limited reference data. MSEM offers competitive performance, particularly when texture information is minimal.

Area of Science:

  • Computer Vision
  • Machine Learning
  • Remote Sensing

Background:

  • Real-world image classification often suffers from ill-posedness due to scarce reference samples and spatial autocorrelation.
  • This is particularly prevalent in remote sensing (RS) image mapping tasks.

Purpose of the Study:

  • To introduce an original inductive learning multiscale image classifier, Multiscale Semisupervised Expectation Maximization (MSEM), designed for image classification with limited reference data.
  • To evaluate MSEM's performance against established and novel classifiers in challenging RS image classification scenarios.

Main Methods:

  • MSEM combines the multiscale modified Pappas adaptive clustering (MPAC) algorithm with a sample-based semisupervised expectation maximization (SEM) classifier.
  • The proposed classifier was benchmarked against MPAC, SEM, contextual SEM (CSEM), and standard classifiers on two RS datasets with limited reference samples.

Related Experiment Videos

Main Results:

  • MSEM demonstrated competitive overall image mapping performance compared to other classifiers.
  • Quantitative map quality indexes supported theoretical considerations and expert evaluations, despite weak subjective results.
  • MSEM incurred a computational overhead 3-6 times greater than its closest rival, SEM.

Conclusions:

  • Semisupervised classifiers utilizing Gaussian mixture models are effective when reference data is scarce and texture information is negligible.
  • MSEM shows promise for ill-posed image classification tasks, especially in remote sensing, despite its higher computational cost.