Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Search research articles

Related Experiment Videos

MAGNet: enhancing action recognition with multimodal fusion and adaptive graph convolution.

Hui Dong¹, Fengqing Zhang², Chunli Liu³

¹School of Communication, WeiFang University, Weifang, 261061, China. dhcm@wfu.edu.cn.

Scientific Reports

|March 27, 2026

Summary

This summary is machine-generated.

Related Concept Videos

Association Areas of the Cortex

Association Areas of the Cortex

Association areas are regions of the cerebral cortex that do not have a specific sensory or motor function. Instead, they integrate and interpret information from various sources to enable higher cognitive processes such as memory, learning, and decision-making. Some key association areas include the following:
Prefrontal Association Area: This area is located in the frontal lobe and is involved in planning, decision-making, and moderating social behavior. It connects with primary motor areas,...

Tagging and Fusion Proteins

Tagging and Fusion Proteins

Proteins are involved in several cellular processes and biochemical reactions. Analyzing a specific protein of interest requires it to be isolated from the other proteins in the cell. This is achieved by overexpressing the specific gene in a suitable host to produce large quantities of the target protein. A tag or label is recombined with the gene to produce a fusion protein containing the target protein and the tag. The tags on these fusion proteins can then be used for easy detection and...

You might also read

Related Articles

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

Sort by

Same author

Prospective Association Between Weight Variability and Subsequent Long-Term Weight Loss in the CALERIE Study.

Obesity (Silver Spring, Md.)·2026

Same author

Functional unblinding in a trial of local tranexamic acid injection during septoplasty.

European archives of oto-rhino-laryngology : official journal of the European Federation of Oto-Rhino-Laryngological Societies (EUFOS) : affiliated with the German Society for Oto-Rhino-Laryngology - Head and Neck Surgery·2026

Same author

Interpreting survival differences in head and neck squamous cell carcinoma of unknown primary: time-dependent treatment and biological selection.

European archives of oto-rhino-laryngology : official journal of the European Federation of Oto-Rhino-Laryngological Societies (EUFOS) : affiliated with the German Society for Oto-Rhino-Laryngology - Head and Neck Surgery·2026

Same author

The secretory production of cutinase ICCG-A2 in Corynebacterium glutamicum using acetate as the sole carbon source.

Journal of biotechnology·2026

Same author

Framework Electronegativity Governs Interfacial Transport Kinetics in Lithium-Metal Batteries.

Journal of the American Chemical Society·2026

Same author

Pollution Characteristics, Potential Sources and Transport Pathways of O<sub>3</sub> and VOCs in a Coastal City of East China.

Bulletin of environmental contamination and toxicology·2026

Same journal

MT-MRI for detection of renal interstitial fibrosis in renovascular disease.

Scientific reports·2026

Same journal

Detection of underground objects from GPR data using a lightweight YOLO-based approach.

Scientific reports·2026

Same journal

Early systemic inflammatory-metabolic trajectory phenotypes are associated with survival outcomes in metastatic renal cell carcinoma treated with nivolumab.

Scientific reports·2026

Same journal

Water balance components in a dry-seeded rice-wheat system: Untangling the effects of tillage and mulching practices.

Scientific reports·2026

Same journal

Topological approaches to quantum tensor train compression via ZX-calculus and SVD.

Scientific reports·2026

Same journal

determinants of flood impacts and adaptive capacity among market vendors in Walukuba-Masese, Jinja city, Uganda.

Scientific reports·2026

See all related articles

This study introduces the Multimodal Adaptive Graph Convolutional Network (MAGNet) for robust human action recognition. MAGNet enhances multimodal data fusion and handles missing data, achieving state-of-the-art results on benchmark datasets.

Area of Science:

Computer Vision
Machine Learning
Artificial Intelligence

Background:

Human action recognition is crucial for applications like VR and surveillance.
Current deep learning models struggle with multimodal fusion, missing data, and complex environments.

Purpose of the Study:

To develop a novel human action recognition model that overcomes limitations in multimodal data fusion and robustness.
To introduce a method capable of handling incomplete or noisy data effectively.

Main Methods:

Proposed the Multimodal Adaptive Graph Convolutional Network (MAGNet).
Integrated adaptive graph convolutions and cross-modal self-attention for enhanced fusion.
Utilized a VQ-VAE generative model to handle missing data and generate pose features.

Keywords:

Cross-modal self-attention Graph convolution Human action recognition Multimodal learning VQ-VAE

Related Experiment Videos

Main Results:

Achieved state-of-the-art performance on NTU RGB+D and UTD-MHAD datasets.
Reached 95.2% (XSub) and 98.8% (XView) accuracy on NTU RGB+D.
Demonstrated strong robustness in multimodal fusion and complex scenes (occlusion, lighting variation).

Conclusions:

MAGNet significantly outperforms existing methods in human action recognition.
The model shows high robustness and effectiveness in real-world conditions.
The proposed approach advances multimodal data fusion for action recognition tasks.