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

Semisupervised biased maximum margin analysis for interactive image retrieval.

Lining Zhang1, Lipo Wang, Weisi Lin

  • 1School of Electrical and Electronic Engineering, Nanyang Technological University, Singapore. zhan0327@e.ntu.edu.sg

IEEE Transactions on Image Processing : a Publication of the IEEE Signal Processing Society
|December 14, 2011
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

Symmetric Entropy-Constrained Video Coding for Machines.

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

From Global to Granular: Revealing IQA Model Performance Via Correlation Surface.

IEEE transactions on pattern analysis and machine intelligence·2026
Same author

Decouple-Then-Synergize: A Self-Paced Collaborative Learning Network for RGB-T Snowy Urban Scene Parsing.

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

Research on medical students' medical ethics reasoning ability-analysis based on the intermediate concept measurement method.

BMC medical ethics·2026
Same author

Real-World Nighttime Image Dehazing via Bayesian-Based Fractional-Order Variational Model.

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

LB-PTQ: Effective Low-Bit Post-Training Quantization for Vision Transformers.

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

Change-Prior-Guided Unsupervised Change Detection of Heterogeneous Remote Sensing Images.

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

AgonicDreamer: Enhancing Multi-View Consistency in Text-to-3D Generation via Rectified Score Distillation.

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

BiCM-Prompt: Bidirectional Cross-Modal Prompt Tuning for Class-Incremental Learning on Multisource Remote Sensing Images.

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

GoP-based Quality Enhancement on Video Compression.

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

Align then Tensorize: Multi-Level Consistent Anchor Graph Learning for Scalable Multi-View Clustering.

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

Beyond Fidelity: Diverse Image Synthesis via Retrieval-Augmented Diffusion.

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

This study introduces Biased Maximum Margin Analysis (BMMA) and Semi-Supervised BMMA (SemiBMMA) to improve content-based image retrieval (CBIR) relevance feedback (RF). These methods better utilize user feedback and unlabeled data for more effective image search.

Area of Science:

  • Computer Science
  • Artificial Intelligence
  • Information Retrieval

Background:

  • Content-based image retrieval (CBIR) systems aim to bridge the semantic gap between visual features and user concepts.
  • Relevance feedback (RF) is crucial for enhancing CBIR performance by incorporating user input.
  • Existing Support Vector Machine (SVM)-based RF methods have limitations in handling distinct feedback types and utilizing unlabeled data.

Purpose of the Study:

  • To propose novel methods, Biased Maximum Margin Analysis (BMMA) and Semi-Supervised BMMA (SemiBMMA), to address drawbacks in SVM-based RF for CBIR.
  • To effectively integrate distinct positive and negative feedback properties and leverage unlabeled samples.
  • To improve the accuracy and efficiency of CBIR systems.

Main Methods:

Related Experiment Videos

  • Developed BMMA to differentiate between positive and negative feedback based on local analysis.
  • Introduced SemiBMMA, incorporating a Laplacian regularizer to utilize unlabeled sample information.
  • Formulated the problem as a subspace learning task with an automatic dimensionality determination approach.

    • Main Results:

    • BMMA successfully differentiates feedback types, enhancing classifier construction.
    • SemiBMMA effectively integrates unlabeled data, further improving classifier performance.
    • Experiments on a large real-world image database show significant performance improvements in CBIR systems.

    Conclusions:

    • The proposed BMMA and SemiBMMA methods offer significant advancements for SVM-based RF in CBIR.
    • These approaches effectively address the limitations of existing methods by better utilizing feedback and unlabeled data.
    • The developed techniques lead to substantially improved CBIR system performance.