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

You might also read

Related Articles

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

Sort by

Same author

Modeling takeover decisions in driving automation: a multilevel drift-diffusion model (MDDM) framework integrating human, system, and environmental factors.

Accident; analysis and prevention·2026

Same author

Single-Nucleus RNA Sequencing Reveals Muscle Fiber Cell Heterogeneity During Human Skeletal Muscle Aging.

Aging cell·2026

Same author

Ovarian Reserve and Endothelial Health in Healthy Reproductive Age Women.

The Journal of clinical endocrinology and metabolism·2026

Same author

Optimizing followers' car-following behaviors using rear-facing eHMIs.

Accident; analysis and prevention·2026

Same author

Physical activity levels, awareness, and associations with geriatric syndromes in hospitalized older adults: a cross-sectional study.

BMC geriatrics·2026

Same author

Users' Trust Evolvement in Fully Driverless Robotaxis During First Ride: An On-Road Study.

Human factors·2026

Same journal

Benchmarking the Robustness of Autonomous Driving to Environmental Illusions: A Lane Perception Perspective.

IEEE transactions on pattern analysis and machine intelligence·2026

Same journal

Learning Topology-Aware Representations via Test-Time Adaptation for Anomaly Segmentation.

IEEE transactions on pattern analysis and machine intelligence·2026

Same journal

TraGraph-GS: Trajectory Graph-based Gaussian Splatting for Arbitrary Large-Scale Scene Rendering.

IEEE transactions on pattern analysis and machine intelligence·2026

Same journal

SWIFT: A Small-World Interaction Framework for Flow-Aware Trajectory Prediction in Autonomous Driving.

IEEE transactions on pattern analysis and machine intelligence·2026

Same journal

HardFlow: Hard-Constrained Sampling for Flow-Matching Models Via Trajectory Optimization.

IEEE transactions on pattern analysis and machine intelligence·2026

Same journal

Industrial Brain: Self-Evolving Neuro-Symbolic Autonomy with Causal Resilience for Cyber-Physical Systems.

IEEE transactions on pattern analysis and machine intelligence·2026

See all related articles

Search research articles

Related Experiment Video

Updated: May 24, 2025

Multifunctional Setup for Studying Human Motor Control Using Transcranial Magnetic Stimulation, Electromyography, Motion Capture, and Virtual Reality

Multifunctional Setup for Studying Human Motor Control Using Transcranial Magnetic Stimulation, Electromyography, Motion Capture, and Virtual Reality

Published on: September 3, 2015

PhysMLE: Generalizable and Priors-Inclusive Multi-Task Remote Physiological Measurement.

Jiyao Wang, Hao Lu, Ange Wang

IEEE Transactions on Pattern Analysis and Machine Intelligence

|March 3, 2025

Summary

This summary is machine-generated.

This study introduces PhysMLE, a novel method for multi-task remote physiological measurement using domain generalization. PhysMLE effectively measures multiple vital signs from face videos, improving generalizability across diverse datasets.

More Related Videos

Simultaneous Recordings of Cortical Local Field Potentials, Electrocardiogram, Electromyogram, and Breathing Rhythm from a Freely Moving Rat

Simultaneous Recordings of Cortical Local Field Potentials, Electrocardiogram, Electromyogram, and Breathing Rhythm from a Freely Moving Rat

Published on: April 2, 2018

Simultaneous Scalp Electroencephalography EEG, Electromyography EMG, and Whole-body Segmental Inertial Recording for Multi-modal Neural Decoding

Simultaneous Scalp Electroencephalography EEG, Electromyography EMG, and Whole-body Segmental Inertial Recording for Multi-modal Neural Decoding

Published on: July 26, 2013

Related Experiment Videos

Last Updated: May 24, 2025

Multifunctional Setup for Studying Human Motor Control Using Transcranial Magnetic Stimulation, Electromyography, Motion Capture, and Virtual Reality

Multifunctional Setup for Studying Human Motor Control Using Transcranial Magnetic Stimulation, Electromyography, Motion Capture, and Virtual Reality

Published on: September 3, 2015

Simultaneous Recordings of Cortical Local Field Potentials, Electrocardiogram, Electromyogram, and Breathing Rhythm from a Freely Moving Rat

Simultaneous Recordings of Cortical Local Field Potentials, Electrocardiogram, Electromyogram, and Breathing Rhythm from a Freely Moving Rat

Published on: April 2, 2018

Simultaneous Scalp Electroencephalography EEG, Electromyography EMG, and Whole-body Segmental Inertial Recording for Multi-modal Neural Decoding

Simultaneous Scalp Electroencephalography EEG, Electromyography EMG, and Whole-body Segmental Inertial Recording for Multi-modal Neural Decoding

Published on: July 26, 2013

Area of Science:

Biomedical Engineering
Computer Science
Physiological Signal Processing

Background:

Remote photoplethysmography (rPPG) measures heart rate from face videos, with domain generalization (DG) crucial for algorithm robustness.
Extending rPPG to multiple vital signs (e.g., respiration, blood oxygen saturation) presents challenges in achieving generalizability due to sparse and imbalanced label spaces.
Existing multi-task learning approaches can suffer from the seesaw effect, hindering task-specific feature learning.

Purpose of the Study:

To develop an effective and generalizable model for multi-task remote physiological measurement.
To address the challenges of sparse and imbalanced label spaces in multi-task rPPG.
To introduce a novel framework that leverages shared and task-specific features for improved multi-modal physiological signal estimation.

Main Methods:

Designed an end-to-end Mixture of Low-rank Experts for multi-task remote Physiological measurement (PhysMLE) model.
Utilized a novel router mechanism within PhysMLE to manage task specifications and correlations.
Incorporated physiological prior knowledge to mitigate label space imbalance and enhance multi-task learning.

Main Results:

PhysMLE demonstrated effectiveness and efficiency in extensive experiments on the proposed Multi-Source Synsemantic Domain Generalization (MSSDG) benchmark and intra-dataset evaluations.
The model successfully handles the complexities of multi-task learning for remote physiological measurements.
Achieved improved generalizability for simultaneous estimation of multiple vital signs.

Conclusions:

PhysMLE offers a robust solution for multi-task remote physiological measurement, outperforming existing methods in generalizability.
The proposed MSSDG protocol and accompanying dataset provide valuable resources for future research in this domain.
The study highlights the potential of low-rank expert models and physiological priors for advancing multi-modal rPPG applications.