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

Performance measures for video object segmentation and tracking.

Ciğdem Eroğlu Erdem1, Bülent Sankur, A Murat Tekalp

  • 1Momentum Digital Media Technology, Inc, Istanbul 80815, Turkey. cigdem@ieee.org

IEEE Transactions on Image Processing : a Publication of the IEEE Signal Processing Society
|January 15, 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

Padé Neurons for Efficient Neural Models.

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2026
Same author

Improved drowsiness detection in drivers through optimum pairing of EEG features using an optimal EEG channel comparable to a multichannel EEG system.

Medical & biological engineering & computing·2025
Same author

Association of Visual-Based Signals with Electroencephalography Patterns in Enhancing the Drowsiness Detection in Drivers with Obstructive Sleep Apnea.

Sensors (Basel, Switzerland)·2024
Same author

Do Thickening Agents Used in Dysphagia Diet Affect Drug Bioavailability?

European journal of pharmaceutical sciences : official journal of the European Federation for Pharmaceutical Sciences·2022
Same author

End-to-End Rate-Distortion Optimized Learned Hierarchical Bi-Directional Video Compression.

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2021
Same author

Realizing a Low-Power Head-Mounted Phase-Only Holographic Display by Light-Weight Compression.

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2020
Same journal

Change-Prior-Guided Unsupervised Change Detection of Heterogeneous Remote Sensing Images.

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

We developed new quantitative measures to assess video object segmentation and tracking performance without needing ground-truth data. These methods analyze spatial and temporal differences to evaluate segmentation accuracy and track reliability.

Area of Science:

  • Computer Vision
  • Image Processing
  • Machine Learning

Background:

  • Evaluating video object segmentation and tracking performance typically requires ground-truth (GT) segmentation maps, which are often unavailable or costly to create.
  • Existing methods lack robust quantitative measures for assessing performance in scenarios devoid of GT data.

Purpose of the Study:

  • To propose novel, quantitative performance measures for video object segmentation and tracking that do not rely on ground-truth segmentation maps.
  • To enable objective evaluation and localization of errors in segmentation and tracking algorithms.

Main Methods:

  • The proposed measures utilize spatial differences in color and motion along object boundaries.
  • Temporal differences in color histograms between consecutive frames are also employed.

Related Experiment Videos

  • Canonical correlation analysis was used to validate the proposed measures against GT-based metrics.

    • Main Results:

    • The developed measures can effectively localize regions with good or poor segmentation and tracking performance, both spatially and temporally.
    • A single numerical score can be derived by combining the measures to represent overall segmentation and tracking quality.
    • Validation confirmed a strong correlation between the proposed GT-free measures and traditional GT-based metrics.

    Conclusions:

    • The proposed quantitative measures offer a viable and effective solution for evaluating video object segmentation and tracking without ground-truth data.
    • These measures facilitate more accessible and widespread performance assessment of computer vision algorithms.
    • The findings enable better understanding and improvement of segmentation and tracking algorithms in real-world applications.