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

Multilevel depth and image fusion for human activity detection.

Bingbing Ni, Yong Pei, Pierre Moulin

    IEEE Transactions on Cybernetics
    |September 3, 2013
    PubMed
    Summary

    This study introduces a new framework for recognizing complex human activities by fusing grayscale and depth camera data. This multi-level fusion significantly improves activity recognition and localization accuracy compared to 2D image-based methods.

    Related Concept Videos

    Depth Perception and Spatial Vision01:15

    Depth Perception and Spatial Vision

    Depth perception is the ability to perceive objects three-dimensionally. It relies on two types of cues: binocular and monocular. Binocular cues depend on the combination of images from both eyes and how the eyes work together. Since the eyes are in slightly different positions, each eye captures a slightly different image. This disparity between images, known as binocular disparity, helps the brain interpret depth. When the brain compares these images, it determines the distance to an object.

    You might also read

    Related Articles

    Articles linked to this work by shared authors, journal, and citation graph.

    Sort by
    Same author

    Atmospheric CO₂-to-acetaldehyde artificial photosynthesis in metallo hydrogen-bonded organic frameworks.

    Nature communications·2026
    Same author

    Learning Gaze Synthesizer via 3D-eye Controlled Diffusion and Cross-domain Feature Alignment.

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

    Semantic Frame Interpolation.

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

    Piclidenoson, an A3 adenosine receptor agonist, demonstrates clinical benefit in canine osteoarthritis: a pilot study.

    Frontiers in veterinary science·2026
    Same author

    Suppressing Polysulfide Shuttling and Enhancing Reaction Kinetics of Lithium-Sulfur Batteries through the Architecture of a 3D TiO<sub>2</sub>@CMK-3 Composite Cathode.

    ACS applied materials & interfaces·2026
    Same author

    Theoretical design of diatomic catalysts for selective hydrodeoxygenation of furfural to 2-methylfuran.

    Physical chemistry chemical physics : PCCP·2026

    Area of Science:

    • Computer Vision
    • Artificial Intelligence
    • Human-Computer Interaction

    Background:

    • Current 2D image-based activity recognition methods struggle with unreliable feature detection and inaccurate interaction modeling.
    • Limitations stem from relying solely on visual features without depth information.

    Purpose of the Study:

    • To propose a novel framework for complex human activity recognition and localization.
    • To effectively fuse grayscale and depth image data at multiple processing levels.
    • To enhance the accuracy of detecting salient visual features and modeling interaction contexts.

    Main Methods:

    • Developed a multi-level fusion framework integrating conventional camera (grayscale) and depth sensor data.
    • Applied depth-based filters for improved individual visual feature detection (e.g., human/object detection).

    Related Experiment Videos

  • Extracted 3D spatial-temporal contexts using fused grayscale and depth information for interaction modeling.
  • Utilized depth data for indoor scene type differentiation.
  • Integrated multi-level information using a latent structural model for activity detection.

    • Main Results:

    • Demonstrated superior performance on activity recognition benchmarks, including one with depth data.
    • Achieved higher recognition and localization accuracies compared to previous methods.
    • Validated effectiveness on a challenging database featuring complex human-human, human-object, and human-surroundings interactions.

    Conclusions:

    • The proposed multi-level grayscale + depth fusion scheme significantly enhances complex human activity recognition and localization.
    • Integrating depth information addresses limitations of 2D-only approaches, leading to more reliable feature detection and context modeling.
    • This framework offers a robust solution for advanced activity understanding in various interactive scenarios.