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

Segmentation framework based on label field fusion.

Pierre-Marc Jodoin1, Max Mignotte, Christophe Rosenberger

  • 1Département d'informatique, Université de Sherbrooke, Sherbrooke QC J1K 2R1, Canada. pierre-marc.jodoin@usherbrooke.ca

IEEE Transactions on Image Processing : a Publication of the IEEE Signal Processing Society
|October 12, 2007
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

Reinforcement Learning for Unsupervised Domain Adaptation in Spatio-Temporal Echocardiography Segmentation.

IEEE transactions on medical imaging·2026
Same author

BundleParc: Consistent white matter bundle parcellation without tractography.

Medical image analysis·2026
Same author

Challenges and best practices when using ComBAT to harmonize diffusion MRI data.

Scientific reports·2025
Same author

Estimation of Segmental Longitudinal Strain in Transesophageal Echocardiography by Deep Learning.

IEEE journal of biomedical and health informatics·2025
Same author

Fusing Echocardiography Images and Medical Records for Continuous Patient Stratification.

IEEE transactions on ultrasonics, ferroelectrics, and frequency control·2025
Same author

Exploring the robustness of TractOracle methods in RL-based tractography.

Medical image analysis·2025
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
Same journal

Multi-Branch Tree-based Fusion Neural Architecture Search with Zero-Cost Screen for Multi-Modal Classification.

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

This study introduces a new fusion framework that combines label fields for computer vision tasks. The method offers conceptual simplicity and efficient parallel implementation for applications like motion segmentation.

Area of Science:

  • Computer Vision
  • Image Processing
  • Machine Learning

Background:

  • Traditional fusion methods often rely on raw observation data.
  • Segmentation maps and object shape information are crucial for scene understanding.
  • Efficient algorithms are needed for real-time computer vision applications.

Purpose of the Study:

  • To propose a novel fusion framework utilizing label fields instead of observation data.
  • To develop a method that fuses segmentation maps and spatial region maps.
  • To enable robust performance in various computer vision applications.

Main Methods:

  • A novel fusion framework is presented, taking two label fields as input: a segmentation map and a spatial region map.
  • Fusion is achieved using a global energy function minimized by a deterministic iterative conditional mode (ICM) algorithm.

Related Experiment Videos

  • The energy function can support pure fusion or a fusion-reaction strategy, with a data-related term for well-posed optimization.

    • Main Results:

    • The proposed framework demonstrates conceptual simplicity and requires a small number of parameters.
    • The deterministic ICM algorithm allows for a simple and fast optimization process.
    • The approach is suitable for natural implementation on parallel architectures, enhancing efficiency.

    Conclusions:

    • The novel fusion framework offers an effective alternative to traditional data-driven fusion methods.
    • Its efficiency, simplicity, and parallelizability make it advantageous for computer vision tasks.
    • The framework is versatile and applicable to motion segmentation, motion estimation, and occlusion detection.