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

Mean shift is a bound optimization.

M Fashing, C Tomasi

    IEEE Transactions on Pattern Analysis and Machine Intelligence
    |March 8, 2005
    PubMed
    Summary
    This summary is machine-generated.

    Mean shift is demonstrated to be equivalent to Newton's method for piecewise constant kernels. This optimization procedure is proven to be a quadratic bound maximization for all kernel types.

    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

    Social Inequalities in the Association between Diabetes and Complete Edentulism.

    Journal of dental research·2025
    Same author

    Sexual quality of life in patients with pemphigus: A case-control study.

    Journal of the European Academy of Dermatology and Venereology : JEADV·2023
    Same author

    A randomized split-face clinical trial of conventional vs indoor-daylight photodynamic therapy for the treatment of multiple actinic keratosis of the face and scalp and photoaging.

    The Journal of dermatological treatment·2021
    Same author

    Clinical and dermoscopic changes of acquired melanocytic nevi of patients treated with afamelanotide.

    Photochemical & photobiological sciences : Official journal of the European Photochemistry Association and the European Society for Photobiology·2021
    Same author

    Occupational low-level exposure to hard metals: cobalt and tungsten biomonitoring as an effective tool to evaluate the effectiveness of industrial hygiene interventions for risk management.

    Biomarkers : biochemical indicators of exposure, response, and susceptibility to chemicals·2020
    Same author

    Effects of MAL-PDT, ingenol mebutate and diclofenac plus hyaluronate gel monitored by high-frequency ultrasound and digital dermoscopy in actinic keratosis - a randomized trial.

    Journal of the European Academy of Dermatology and Venereology : JEADV·2019
    Same journal

    TraGraph-GS: Trajectory Graph-based Gaussian Splatting for Arbitrary Large-Scale Scene Rendering.

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

    SWIFT: A Small-World Interaction Framework for Flow-Aware Trajectory Prediction in Autonomous Driving.

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

    HardFlow: Hard-Constrained Sampling for Flow-Matching Models Via Trajectory Optimization.

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

    Industrial Brain: Self-Evolving Neuro-Symbolic Autonomy with Causal Resilience for Cyber-Physical Systems.

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

    Adaptive Hardness-Driven Dictionary Distillation for Incomplete Streaming View Clustering.

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

    Mixture of Global and Local Experts with Diffusion Transformer for Controllable Face Generation.

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

    Area of Science:

    • Machine Learning
    • Optimization Algorithms
    • Computer Vision

    Background:

    • Mean shift is a widely used non-parametric clustering and mode-seeking algorithm.
    • Its application as an optimization procedure is an area of ongoing research.
    • Understanding its theoretical underpinnings enhances its applicability.

    Purpose of the Study:

    • To deepen the theoretical understanding of the mean shift algorithm.
    • To establish a connection between mean shift and established optimization techniques.
    • To prove the general mathematical property of the mean shift procedure.

    Main Methods:

    • Theoretical analysis of the mean shift algorithm.
    • Mathematical derivation and proof.
    • Comparison with Newton's method for specific kernel types.

    Related Experiment Videos

    Main Results:

    • Demonstration of mean shift's equivalence to Newton's method for piecewise constant kernels.
    • Proof that mean shift acts as a quadratic bound maximization for all kernels.
    • Provides a new perspective on mean shift as a general optimization tool.

    Conclusions:

    • Mean shift is a robust optimization procedure with theoretical links to Newton's method.
    • The quadratic bound maximization property offers insights into its convergence and behavior.
    • This work contributes to a more comprehensive understanding of mean shift in machine learning and optimization.