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

Related Concept Videos

Fixed Action Patterns

Fixed Action Patterns

A fixed action pattern (FAP) is a specific, hard-wired sequence of behaviors that occurs in response to an external stimulus, called a sign stimulus. The behavior is “fixed” because it is essentially unchangeable—proceeding similarly across individuals of a species every time it occurs.

Functional Classification of Joints

Functional Classification of Joints

Functional Classification of Joints
The functional classification of joints is determined by the amount of mobility between the adjacent bones. Joints are functionally classified as a synarthrosis or immobile joint, an amphiarthrosis or slightly moveable joint, or as a diarthrosis, a freely moveable joint. Fibrous and cartilaginous joints can be functionally classified as either synarthroses or amphiarthroses, whereas all synovial joints are classified as diarthroses.
Synarthrosis
An immobile...

Force Classification

Force Classification

Forces play a crucial role in the study of physics and engineering. They are essential in describing the motion, behavior, and equilibrium of objects in the physical world. Forces can be classified based on their origin, type, and direction of action.
Contact and non-contact forces are two of the most widely used categories of forces. As the name suggests, contact forces require physical contact between two objects to act upon each other. Examples of contact forces include frictional,...

Light Acquisition

Light Acquisition

In order to produce glucose, plants need to capture sufficient light energy. Many modern plants have evolved leaves specialized for light acquisition. Leaves can be only millimeters in width or tens of meters wide, depending on the environment. Due to competition for sunlight, evolution has driven the evolution of increasingly larger leaves and taller plants, to avoid shading by their neighbors with contaminant elaboration of root architecture and mechanisms to transport water and nutrients.

You might also read

Related Articles

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

Sort by

Same author

Motion prediction for leader manipulator of teleoperation system with large time delay based on inverse optimal control.

ISA transactions·2026

Same author

Dual-Modal Safety Framework for Robotic-Assisted Bronchoscopy via Endoscopic Vision and Haptic Feedback.

The international journal of medical robotics + computer assisted surgery : MRCAS·2026

Same author

Towards Interpretable Seizure Detection: An Excitation/Inhibition Dynamic Polynomial Network Framework for Electroencephalography.

Sensors (Basel, Switzerland)·2026

Same author

Human-in-the-Loop Control Framework for Robot-Mediated Error Augmentation Training Based on Muscle Synergy Assessment.

IEEE transactions on neural systems and rehabilitation engineering : a publication of the IEEE Engineering in Medicine and Biology Society·2026

Same author

Design, Analysis, and Characterization of a Small-Scale High-Torque Magnetorheological Brake for Haptic Applications.

IEEE transactions on haptics·2026

Same author

Robust Human-to-Robot Handover System Under Adverse Lighting.

Biomimetics (Basel, Switzerland)·2026

Search research articles

Related Experiment Videos

Cross-Modal Feature Adapter for Few-Shot Human Activity Recognition.

Xin Liu, Lei Zhang, Wenbo Huang

IEEE Journal of Biomedical and Health Informatics

|July 7, 2026

Summary

This study introduces a novel cross-modal data augmentation technique to address the challenge of limited labeled data in human activity recognition (HAR). The method effectively leverages image data to improve deep learning model performance in few-shot HAR scenarios.

Related Experiment Videos

Area of Science:

Computer Science
Artificial Intelligence
Machine Learning

Background:

Deep learning has shown great success in sensor-based human activity recognition (HAR).
A major challenge in HAR is the scarcity of labeled sensor data, hindering model development, especially in few-shot learning scenarios.
This limitation requires extensive manual annotation, which is time-consuming and labor-intensive.

Purpose of the Study:

To address the data scarcity issue in few-shot human activity recognition.
To develop a cross-modal data augmentation strategy that utilizes readily available image data.
To improve the efficiency and effectiveness of deep learning models in HAR with limited labeled sensor data.

Main Methods:

A cross-modal data augmentation approach is proposed, using activity label text to retrieve relevant images and create an augmented dataset.
A novel cross-modal feature adapter is designed to fine-tune a pre-trained CLIP image encoder, aligning image-sensor data pairs.
A learnable residual ratio adaptively blends knowledge from the original CLIP model and few-shot training samples for faster convergence and a streamlined sensor encoder.

Main Results:

The proposed method significantly outperforms existing state-of-the-art HAR baselines across all tested few-shot scenarios.
Experiments conducted on three public HAR benchmarks validate the effectiveness of the cross-modal approach.
The method demonstrates faster training convergence and achieves superior performance compared to traditional methods.

Conclusions:

The developed cross-modal data augmentation and feature adaptation method effectively overcomes the challenge of limited labeled data in few-shot HAR.
The approach offers a promising solution for enhancing the practical application of deep learning in HAR, particularly in data-scarce environments.
The study highlights the potential of leveraging multi-modal data, specifically images, to boost the performance of sensor-based activity recognition systems.