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

Learning a Deep Model for Human Action Recognition from Novel Viewpoints.

Hossein Rahmani, Ajmal Mian, Mubarak Shah

    IEEE Transactions on Pattern Analysis and Machine Intelligence
    |April 15, 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

    Yield stress fluids in microfluidics: research, applications and opportunities.

    Lab on a chip·2026
    Same author

    DynaPURLS: Dynamic Refinement of Part-Aware Representations for Skeleton-Based Zero-Shot Action Recognition.

    IEEE transactions on pattern analysis and machine intelligence·2026
    Same author

    Second-Order Robust Iterative Pose Optimization for Fine-Grained Cross-View Localization.

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

    STROKEVISION-BENCH: A MULTIMODAL VIDEO AND 2D POSE BENCHMARK FOR TRACKING STROKE RECOVERY.

    IEEE International Workshop on Machine Learning for Signal Processing : [proceedings]. IEEE International Workshop on Machine Learning for Signal Processing·2026
    Same author

    Two-phase simulations of viscoplastic flow in superhydrophobic microchannels: interface stability, plug dynamics, and drag reduction.

    Lab on a chip·2026
    Same author

    Diffusion-Driven Self-Supervised Learning for Shape Reconstruction and Pose Estimation.

    IEEE transactions on pattern analysis and machine intelligence·2025
    Same journal

    Relation DETR+: Exploring Explicit Position Relation Prior for Dense Prediction.

    IEEE transactions on pattern analysis and machine intelligence·2026
    Same journal

    RBF++: Quantifying and Optimizing Reasoning Boundaries across Measurable and Unmeasurable Capabilities for Chain-of-Thought Reasoning.

    IEEE transactions on pattern analysis and machine intelligence·2026
    Same journal

    CAFE: Cross-View Adaptive Fusion and Cluster Center Enhancement for Robust Multi-View Clustering.

    IEEE transactions on pattern analysis and machine intelligence·2026
    Same journal

    DIVER: Reinforced Diffusion Breaks Imitation Bottlenecks in End-to-End Autonomous Driving.

    IEEE transactions on pattern analysis and machine intelligence·2026
    Same journal

    Ethics-Aware Safe Reinforcement Learning for Rare-Event Risk Control in Interactive Urban Driving.

    IEEE transactions on pattern analysis and machine intelligence·2026
    Same journal

    Learning Shape Anchors for Holistic Indoor Scene Understanding.

    IEEE transactions on pattern analysis and machine intelligence·2026
    See all related articles

    This study introduces a Robust Non-Linear Knowledge Transfer Model (R-NKTM) for recognizing human actions from new viewpoints. The R-NKTM effectively transfers knowledge across views without retraining, outperforming current methods on benchmark datasets.

    Area of Science:

    • Computer Vision
    • Machine Learning
    • Human-Computer Interaction

    Background:

    • Recognizing human actions from unseen viewpoints presents significant challenges in computer vision.
    • Existing methods often require retraining or fine-tuning for new views or action classes.

    Purpose of the Study:

    • To develop a novel model for robust human action recognition from unknown views.
    • To enable efficient knowledge transfer across different viewpoints without requiring view-specific training data.

    Main Methods:

    • Proposed a Robust Non-Linear Knowledge Transfer Model (R-NKTM), a deep fully-connected neural network.
    • Learned non-linear transformations to map actions from unknown views to a shared virtual view.
    • Trained the model using 2D projections of synthetic 3D human models and real motion capture data.

    Related Experiment Videos

    Main Results:

    • The R-NKTM successfully generalizes to real-world human action videos from novel viewpoints.
    • A single R-NKTM model handles all actions and viewpoints without re-training, demonstrating scalability.
    • Achieved state-of-the-art performance on three benchmark cross-view human action recognition datasets.

    Conclusions:

    • The R-NKTM offers an efficient and effective solution for cross-view human action recognition.
    • The model's ability to learn without explicit viewpoint information simplifies its application.
    • This approach advances the field by enabling robust action recognition across diverse and previously unseen perspectives.