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 Video

Updated: Sep 20, 2025

Photorealistic Learned Landscapes for Augmented Reality
06:54

Photorealistic Learned Landscapes for Augmented Reality

Published on: June 27, 2025

210

Multi-View Large Reconstruction Model via Geometry-Aware Positional Encoding and Attention.

Mengfei Li, Xiaoxiao Long, Yixun Liang

    IEEE Transactions on Visualization and Computer Graphics
    |May 23, 2025
    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

    Integrating network pharmacology and in vivo evaluation reveals chicory (Cichorium intybus L.) extract activates autophagy to alleviate fatty liver hemorrhagic syndrome in laying hens.

    Poultry science·2026
    Same author

    Multiomics Reveal Low Temperatures Play Positive Roles in Flavonoids and Phenolic Accumulation in Meconopsis punicea.

    Physiologia plantarum·2026
    Same author

    Discovery of R-185 as a Potent and Orally Efficacious GPR52 Agonist with Potential for Treating Schizophrenia and Related Neuropsychiatric Disorders.

    Journal of medicinal chemistry·2026
    Same author

    Author Correction: Inactivating SnRK1β1A promotes broad-spectrum disease resistance in rice.

    Nature·2026
    Same author

    Effects of Supplementation with Dimethylglycine Sodium Salt on Immunity, Intestinal Tissue Morphology, and Antioxidant Function in IUGR Lambs.

    Animals : an open access journal from MDPI·2026
    Same author

    Soil pH Amelioration Fosters Persistent Carbon Sinks Through Mineral Stabilization and Aggregate Protection.

    Global change biology·2026
    Same journal

    MesoSplats: Texture Synthesis with Gaussian Splatting.

    IEEE transactions on visualization and computer graphics·2026
    Same journal

    GLLA: A Unified Force-Directed Graph Layout Framework Supporting Local Adjustments.

    IEEE transactions on visualization and computer graphics·2026
    Same journal

    Multi-Perception Crowd: Learning to combine entity and implicit perception for diverse crowd simulation.

    IEEE transactions on visualization and computer graphics·2026
    Same journal

    Hiding in Plain Sight: Camouflaging Real-world Objects.

    IEEE transactions on visualization and computer graphics·2026
    Same journal

    RTF2Mesh: Restricted Tangent Face Based Mesh Compression With Neural Displacement Fields.

    IEEE transactions on visualization and computer graphics·2026
    Same journal

    Practical Occluder Generation for Mobile Games.

    IEEE transactions on visualization and computer graphics·2026
    See all related articles

    The Multi-view Large Reconstruction Model (M-LRM) enhances 3D shape reconstruction from multiple images. This new approach improves geometric quality and training speed compared to previous methods.

    Area of Science:

    • Computer Vision
    • 3D Reconstruction
    • Machine Learning

    Background:

    • Large Reconstruction Models (LRMs) show promise for single-image 3D reconstruction.
    • Extending LRMs to multi-view inputs reveals inefficiencies in quality and convergence.
    • Existing methods often treat multi-view reconstruction as simple image-to-3D translation, neglecting 3D coherence.

    Purpose of the Study:

    • To develop a Multi-view Large Reconstruction Model (M-LRM) for high-fidelity 3D shape generation from multi-view images.
    • To address the limitations of existing LRMs in handling multi-view inputs.
    • To enable 3D-aware reconstruction by leveraging multi-view consistency.

    Main Methods:

    • Introduction of a multi-view consistent cross-attention mechanism for precise information querying across images.

    More Related Videos

    Author Spotlight: Revolutionizing Remote Surgery with Augmented Reality and Robotics for Enhanced Precision and Accessibility
    07:46

    Author Spotlight: Revolutionizing Remote Surgery with Augmented Reality and Robotics for Enhanced Precision and Accessibility

    Published on: August 9, 2024

    859

    Related Experiment Videos

    Last Updated: Sep 20, 2025

    Photorealistic Learned Landscapes for Augmented Reality
    06:54

    Photorealistic Learned Landscapes for Augmented Reality

    Published on: June 27, 2025

    210
    Author Spotlight: Revolutionizing Remote Surgery with Augmented Reality and Robotics for Enhanced Precision and Accessibility
    07:46

    Author Spotlight: Revolutionizing Remote Surgery with Augmented Reality and Robotics for Enhanced Precision and Accessibility

    Published on: August 9, 2024

    859
  • Utilization of 3D priors from input multi-view images to initialize triplane tokens.
  • Formulation of a 3D-aware reconstruction process, moving beyond naive image-to-3D translation.

    • Main Results:

    • M-LRM generates 3D shapes with significantly improved fidelity compared to prior methods.
    • The proposed model demonstrates faster training convergence.
    • Experimental studies confirm substantial performance gains in multi-view 3D reconstruction.

    Conclusions:

    • M-LRM effectively reconstructs high-quality 3D shapes from multi-view images.
    • The 3D-aware approach and multi-view consistent cross-attention are key to M-LRM's success.
    • M-LRM offers a more efficient and effective solution for multi-view 3D reconstruction tasks.