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

Beyond Single Descriptors: Complementary Feature Learning for Image Matching.

Xianguo Yu1,2, Yulong Feng3, Xi Li3

  • 1College of Intelligence Science and Technology, National University of Defense Technology, Changsha 410073, China.

Journal of Imaging
|May 26, 2026
PubMed
Summary
This summary is machine-generated.

Related Concept Videos

Sign Test for Matched Pairs01:17

Sign Test for Matched Pairs

The sign test for matched pairs offers a robust method for comparing two paired samples, often for the effects of an intervention in one of them. This method is very useful in situations where the underlying distribution of the data is unknown. The test compares two related samples—often pre- and post-treatment measurements on the same subjects—to determine if there are significant differences in their median values.
To conduct the sign test, we first calculate the differences in value between...

You might also read

Related Articles

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

Sort by
Same author

NF90 is a novel influenza A virus NS1-interacting protein that antagonizes the inhibitory role of NS1 on PKR phosphorylation.

FEBS letters·2016
Same author

Induction of HOXA9 expression in three-dimensional organotypic culture of the Claudin-low breast cancer cells.

Oncotarget·2016
Same author

Genetic and Clinical Predictive Factors of Sulfonylurea Failure in Patients with Type 2 Diabetes.

Diabetes technology & therapeutics·2016
Same author

Distinct roles of parafibromin in the extracellular environment, cytoplasm and nucleus of osteosarcoma cells.

American journal of translational research·2016
Same author

Survival Predictors for Severe ARDS Patients Treated with Extracorporeal Membrane Oxygenation: A Retrospective Study in China.

PloS one·2016
Same author

A family with Robertsonian translocation: a potential mechanism of speciation in humans.

Molecular cytogenetics·2016
Same journal

Human-AI Interaction in Interventional Radiology: A Narrative Review of Current Applications, Challenges, and Future Directions.

Journal of imaging·2026
Same journal

Coronary Artery Anomalies and Anatomical Variants: Cross-Sectional Diagnostic Imaging and Clinical Background.

Journal of imaging·2026
Same journal

YoLeTooth: A Unified Framework for Joint Tooth Segmentation and Periapical Lesion Detection in Panoramic Radiographs.

Journal of imaging·2026
Same journal

Radiomics-Guided Multi-Sequence Learning for Pathological Complete Response Prediction from Breast MRI with Missing Auxiliary Sequences.

Journal of imaging·2026
Same journal

Cutaneous Thermography in Arthropathies: Quantitative Imaging, Machine Learning, and Clinical Translation.

Journal of imaging·2026
Same journal

Two-Stage Dynamic Synergistic Segmentation Method for Myocardial Pathology.

Journal of imaging·2026
See all related articles

This study introduces a novel complementary local feature description model that simultaneously learns two descriptor maps to enhance feature matching. This approach improves performance in complex visual geometry tasks by overcoming limitations of single descriptor maps.

Area of Science:

  • Computer Vision
  • Machine Learning

Background:

  • Sparse local feature matching is crucial for visual geometry.
  • Improving descriptor discriminative power remains a challenge.
  • Current methods using single descriptor maps struggle with complex scenarios and real-time applications due to resolution reduction and interpolation issues.

Purpose of the Study:

  • To propose an efficient, novel complementary local feature description model.
  • To address the limitations of single descriptor maps in sparse feature matching.
  • To enhance descriptor discriminative power for complex visual geometry tasks.

Main Methods:

  • A single Convolutional Neural Network (CNN) simultaneously learns two descriptor maps.
  • Different loss functions are employed for each descriptor map.
Keywords:
complementary descriptorslocal feature matchingorthogonal losspose estimation

Related Experiment Videos

  • An orthogonal loss function coordinates the learning of the two branches to achieve decoupled and complementary descriptors.

    • Main Results:

    • The proposed model demonstrates superior performance across various visual geometry tasks.
    • Experiments include homography estimation, indoor and outdoor pose estimation, and visual localization.
    • The method effectively overcomes background contamination and enhances descriptor discriminability.

    Conclusions:

    • The novel complementary local feature description model significantly advances sparse feature matching.
    • Simultaneous learning of dual descriptor maps with orthogonal loss improves performance in challenging scenarios.
    • The method offers a promising solution for real-time visual geometry applications.