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

Sparse Bayesian learning for efficient visual tracking.

Oliver Williams1, Andrew Blake, Roberto Cipolla

  • 1Machine Intelligence Laboratory, Department of Engineering, University of Cambridge, Trumpington Street, Cambridge, CB2 1PZ, UK. omcw2@cam.ac.uk

IEEE Transactions on Pattern Analysis and Machine Intelligence
|August 27, 2005
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

From bench to byte: A UK perspective on data-driven cancer research.

European journal of cancer (Oxford, England : 1990)·2026
Same author

Education Research: "Simulation Sharing": Use of a Cardiac Arrest Management Course Developed for Medicine Residents Applied to Neurocritical Care Fellows.

Neurology. Education·2025
Same author

Tumour purity assessment with deep learning in colorectal cancer and impact on molecular analysis.

The Journal of pathology·2024
Same author

Prediction of total and regional body composition from 3D body shape.

NPJ digital medicine·2024
Same author

Near fatal case of mirvetuximab soravtansine-gynx induced interstitial lung disease and a review of the primary literature.

Gynecologic oncology reports·2024
Same author

Efficacy of endoscopic decompression surgery for treatment of lumbar spinal stenosis.

Interventional pain medicine·2024
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
Same journal

Task-KV: Task-aware KV Cache Optimization via Semantic Differentiation of Attention Heads.

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

Achieving Text-based Person Retrieval with Any Granularity.

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

This study introduces a novel object localization method using a probabilistic Relevance Vector Machine (RVM) for improved temporal data fusion. The RVM-based displacement expert enables efficient, real-time object tracking with automatic recovery capabilities.

Area of Science:

  • Computer Vision
  • Machine Learning
  • Statistical Learning

Background:

  • Object localization traditionally uses kernel-Support Vector Machines (SVMs) applied frame-by-frame.
  • Temporal data fusion enhances performance but is challenging to integrate with frame-independent methods.
  • Existing methods often lack automatic initialization and recovery mechanisms.

Purpose of the Study:

  • To develop a robust object localization system leveraging temporal data fusion.
  • To introduce a probabilistic model for enhanced displacement estimation.
  • To enable real-time object tracking with automatic initialization and recovery.

Main Methods:

  • Utilized a fully probabilistic Relevance Vector Machine (RVM) for temporal data fusion.
  • Developed a displacement expert to directly estimate object movement.

Related Experiment Videos

  • Integrated an object detector for verification, initialization, and recovery.
  • Implemented a sparse RVM for efficient real-time processing.

    • Main Results:

    • The RVM approach effectively fuses temporal data for improved object localization.
    • The displacement expert accurately estimates object movement.
    • The system demonstrated real-time tracking performance at frame rate.
    • The method showed viability for long-term region tracking compared to state-of-the-art techniques.

    Conclusions:

    • The proposed RVM-based method offers a viable and efficient solution for real-time object localization and tracking.
    • Temporal fusion via probabilistic models significantly enhances tracking robustness.
    • The system's sparse nature and integrated verification/recovery features make it suitable for demanding applications.