Jove
Visualize
Contact Us

Related Experiment Videos

TMAGIC: A Model-Free 3D Tracker.

Karel Lebeda, Simon Hadfield, Richard Bowden

    IEEE Transactions on Image Processing : a Publication of the IEEE Signal Processing Society
    |April 1, 2017
    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

    Soil organic matter molecular composition with long-term detrital alterations is controlled by site-specific forest properties.

    Global change biology·2022
    Same author

    Weakly Supervised Reinforcement Learning for Autonomous Highway Driving via Virtual Safety Cages.

    Sensors (Basel, Switzerland)·2021
    Same author

    Hollywood 3D: What are the Best 3D Features for Action Recognition?

    International journal of computer vision·2020
    Same author

    Weakly Supervised Learning with Multi-Stream CNN-LSTM-HMMs to Discover Sequential Parallelism in Sign Language Videos.

    IEEE transactions on pattern analysis and machine intelligence·2019
    Same author

    HARD-PnP: PnP Optimization Using a Hybrid Approximate Representation.

    IEEE transactions on pattern analysis and machine intelligence·2018
    Same author

    Texture-Independent Long-Term Tracking Using Virtual Corners.

    IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2015
    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
    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

    This study introduces a new 3D visual tracking method that models 3D motion directly, improving performance on challenging sequences with rapid out-of-plane rotation. This approach enhances object tracking accuracy and bridges visual tracking with 3D reconstruction.

    Area of Science:

    • Computer Vision
    • Robotics
    • 3D Reconstruction

    Background:

    • Current visual tracking methods struggle with rapid out-of-plane rotations and sudden appearance changes.
    • Existing approaches often model 3D motion indirectly through appearance variations, limiting robustness.

    Purpose of the Study:

    • To propose a novel 3D visual tracking framework that directly models 3D motion.
    • To improve tracking accuracy and robustness, especially in scenarios with extreme object rotations.
    • To bridge the gap between visual tracking and 3D structure from motion.

    Main Methods:

    • Modeling 3D motion as explicit 3D motion rather than appearance changes.
    • Developing a new benchmark dataset for extreme out-of-plane rotation sequences.
    • Introducing an online leaderboard to encourage research in 3D tracking.

    Related Experiment Videos

    Main Results:

    • The proposed method successfully tracks sequences with extreme out-of-plane rotations, outperforming 2D trackers.
    • Tracking error was reduced by 46% on challenging benchmark sequences.
    • The approach demonstrates the potential for 3D tracking to support 3D reconstruction.

    Conclusions:

    • Directly modeling 3D motion offers a more general and robust approach to visual tracking.
    • The new benchmark and leaderboard will accelerate research in 3D visual tracking.
    • This work opens new avenues for integrating tracking with 3D reconstruction applications.