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 Video

Updated: May 22, 2026

Design and Analysis for Fall Detection System Simplification

Design and Analysis for Fall Detection System Simplification

Published on: April 6, 2020

Bridging the visual-to-physical gap: physically aligned representations for fall risk analysis.

Xianqi Zhang^1,2, Xingtao Wang¹, Xiaopeng Fan^1,2,3

¹Faculty of Computing, Harbin Institute of Technology, Harbin, China.

Frontiers in Medicine

|May 21, 2026

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

<i>In situ</i> glycosylation-directed H-aggregation of Type I photosensitizers for synergistic biofilm eradication and promoting diabetic wound healing.

Chemical science·2026

Same author

Toward precise management of groundwater by combined heavy metal(loid)s contamination at industrial sites: A machine learning driven source-to-risk zoning framework.

Environmental pollution (Barking, Essex : 1987)·2026

Same author

Mechanical pre-stretch of extracellular matrix regulates podosome dynamics via contact stiffness.

Biophysical journal·2026

Same author

Molecularly Confined Domains Enable Halide-Stable Wide-Bandgap Perovskites.

Advanced materials (Deerfield Beach, Fla.)·2026

Same author

Influence of nonlinear node operating states on the performance of photonic reservoir computing.

Optics express·2026

Same author

Language Supervised Multi-Camera Multi-Object Tracking.

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

Same journal

ABO gene polymorphisms: a molecular bridge linking disease susceptibility to therapeutic outcomes.

Frontiers in medicine·2026

Same journal

Lightweight deep learning framework for intracranial hemorrhage detection in brain CT scans.

Frontiers in medicine·2026

Same journal

Medical education in low- and middle-income countries must be innovated without neglecting its foundational principles.

Frontiers in medicine·2026

Same journal

Development and internal validation of a nomogram for predicting adverse pregnancy outcomes in women with early-onset preeclampsia.

Frontiers in medicine·2026

Same journal

Effectiveness of digital personalized nursing pathway on postoperative rehabilitation in patients with non-small cell lung cancer.

Frontiers in medicine·2026

Same journal

Increased severity and relapse rate in adult anti-nuclear matrix protein 2 antibody positive myositis patients with distal limb weakness: a retrospective study in China.

Frontiers in medicine·2026

See all related articles

PHARL (PHysics-aware Alignment Representation Learning) learns fall representations from video without clinical labels by integrating physics simulation. This approach improves fall analysis and risk prediction, even revealing an emergent severity ordering.

Area of Science:

Computer Vision
Biomechanics
Machine Learning

Background:

Vision-based fall analysis faces challenges inferring physical outcomes from visual data due to subtle differences in mechanics.
Supervised injury prediction is hindered by the difficulty in obtaining reliable clinical labels for falls, often due to video ambiguities and ethical constraints.

Purpose of the Study:

To develop a framework for learning physically meaningful fall representations without requiring clinical outcome labels.
To improve the quality of risk-aligned representations in fall analysis and maintain performance in fall detection.

Main Methods:

Proposed PHARL (PHysics-aware Alignment Representation Learning) framework.
Introduced trajectory-level temporal consistency and physics-aware alignment using simulation-derived contact outcomes to structure the embedding space.

Keywords:

contrastive learning deep learning embedding geometry fall risk analysis representation learning

More Related Videos

A Modified Lean and Release Technique to Emphasize Response Inhibition and Action Selection in Reactive Balance

A Modified Lean and Release Technique to Emphasize Response Inhibition and Action Selection in Reactive Balance

Published on: March 19, 2020

Related Experiment Videos

Last Updated: May 22, 2026

Design and Analysis for Fall Detection System Simplification

Design and Analysis for Fall Detection System Simplification

Published on: April 6, 2020

A Modified Lean and Release Technique to Emphasize Response Inhibition and Action Selection in Reactive Balance

A Modified Lean and Release Technique to Emphasize Response Inhibition and Action Selection in Reactive Balance

Published on: March 19, 2020

Associated video windows with simulation descriptors for capturing impact-relevant dynamics in a feed-forward manner.

Main Results:

PHARL consistently improved risk-aligned representation quality compared to visual-only baselines across four public datasets.
Maintained strong performance on fall detection tasks.
Demonstrated an emergent, interpretable ordinal structure in the learned representation space (Head > Trunk > Supported) without explicit ordinal supervision.

Conclusions:

Physics-aware regularization can induce meaningful structural priors in representation learning for fall analysis.
PHARL offers a promising direction for label-efficient fall analysis by leveraging physics information as a structural regularizer.
The framework successfully addresses the bottleneck of inferring physical outcomes from visual fall data.