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

Distance learning for similarity estimation.

Jie Yu1, Jaume Amores, Nicu Sebe

  • 1Intelligent Systems Group, Kodak Research Labs, Rochester, NY 14615, USA. jerry.j.yu@gmail.com

IEEE Transactions on Pattern Analysis and Machine Intelligence
|January 16, 2008
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

On-Board Object Detection: Multicue, Multimodal, and Multiview Random Forest of Local Experts.

IEEE transactions on cybernetics·2017
Same author

[Role of EuroSCORE and SinoSCORE in prediction of early postoperative quality of life in patients after coronary artery bypass surgery].

Beijing da xue xue bao. Yi xue ban = Journal of Peking University. Health sciences·2015
Same author

Naturally Occurring Stilbenoid TSG Reverses Non-Alcoholic Fatty Liver Diseases via Gut-Liver Axis.

PloS one·2015
Same author

[NT5C2 expression in children with acute leukemia and its clinical significance].

Zhonghua xue ye xue za zhi = Zhonghua xueyexue zazhi·2015
Same author

Tumor necrosis factor-α gene 308G/A polymorphism is not associated with esophageal squamous cell carcinoma risk in Kazakh patients.

International journal of clinical and experimental pathology·2015
Same author

HMGB-1 as a Potential Target for the Treatment of Diabetic Retinopathy.

Medical science monitor : international medical journal of experimental and clinical research·2015
Same journal

Relation DETR+: Exploring Explicit Position Relation Prior for Dense Prediction.

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

RBF++: Quantifying and Optimizing Reasoning Boundaries across Measurable and Unmeasurable Capabilities for Chain-of-Thought Reasoning.

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

CAFE: Cross-View Adaptive Fusion and Cluster Center Enhancement for Robust Multi-View Clustering.

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

DIVER: Reinforced Diffusion Breaks Imitation Bottlenecks in End-to-End Autonomous Driving.

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

Ethics-Aware Safe Reinforcement Learning for Rare-Event Risk Control in Interactive Urban Driving.

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

Learning Shape Anchors for Holistic Indoor Scene Understanding.

IEEE transactions on pattern analysis and machine intelligence·2026
See all related articles

This study introduces novel distance measures for improved similarity estimation, outperforming traditional methods. A boosted framework accounts for heterogeneous data, enhancing accuracy in applications like stereo matching and motion tracking.

Area of Science:

  • Computer Science
  • Data Science
  • Machine Learning

Background:

  • Traditional distance measures like Euclidean and Manhattan often fail with heterogeneous data sources.
  • The assumption of a single isotropic distribution model is frequently inadequate for accurate similarity estimation.

Purpose of the Study:

  • To develop a general guideline for selecting superior distance measures for similarity estimation.
  • To introduce new distance measures derived from harmonic, geometric, and generalized variants using Maximum Likelihood theory.
  • To propose a boosted distance measure framework addressing data heterogeneity for enhanced accuracy.

Main Methods:

  • Statistical analysis of distribution models and distance functions.
  • Derivation of new distance measures based on Maximum Likelihood theory.

Related Experiment Videos

  • Implementation of a boosted distance measure framework to select optimal measures for heterogeneous feature elements.

    • Main Results:

    • New distance measures demonstrate superior feature modeling compared to Euclidean and Manhattan distances.
    • The boosted distance measure framework effectively handles heterogeneous data sources.
    • Robust results were achieved in stereo matching, motion tracking, and image retrieval applications.

    Conclusions:

    • The proposed distance measures and boosted framework offer a more accurate approach to similarity estimation.
    • The methods are effective across diverse applications and benchmark datasets.
    • This work provides a valuable guideline for selecting appropriate distance measures in data analysis.