Meta-rPPG: Remote Heart Rate Estimation Using a Transductive Meta-Learner (2020-07-14T00:00:00.000000Z)

TL;DR

This work proposes a transductive meta-learner that takes unlabeled samples during testing (deployment) for a self-supervised weight adjustment (also known as transductives inference), providing fast adaptation to the distributional changes during deployment.

Abstract

Remote heart rate estimation is the measurement of heart rate without any physical contact with the subject and is accomplished using remote photoplethysmography (rPPG) in this work. rPPG signals are usually collected using a video camera with a limitation of being sensitive to multiple contributing factors, e.g. variation in skin tone, lighting condition and facial structure. End-to-end supervised learning approach performs well when training data is abundant, covering a distribution that doesn't deviate too much from the distribution of testing data or during deployment. To cope with the unforeseeable distributional changes during deployment, we propose a transductive meta-learner that takes unlabeled samples during testing (deployment) for a self-supervised weight adjustment (also known as transductive inference), providing fast adaptation to the distributional changes. Using this approach, we achieve state-of-the-art performance on MAHNOB-HCI and UBFC-rPPG.

Authors

Eugene Lee

1 papers

E. Chen

1 papers

Chen-Yi Lee

1 papers

TL;DR

Abstract

Authors

References68 items

Empirical Bayes Transductive Meta-Learning with Synthetic Gradients

FOAL: Fast Online Adaptive Learning for Cardiac Motion Estimation

PyTorch: An Imperative Style, High-Performance Deep Learning Library

Domain Generalization via Model-Agnostic Learning of Semantic Features

RhythmNet: End-to-End Heart Rate Estimation From Face via Spatial-Temporal Representation

3D Convolutional Neural Networks for Remote Pulse Rate Measurement and Mapping from Facial Video

Remote Heart Rate Measurement From Highly Compressed Facial Videos: An End-to-End Deep Learning Solution With Video Enhancement

Likelihood Ratios for Out-of-Distribution Detection

Deep Compressed Sensing

Remote Photoplethysmograph Signal Measurement from Facial Videos Using Spatio-Temporal Networks

Centralized State Sensing using Sensor Array on Wearable Device

A Meta-Transfer Objective for Learning to Disentangle Causal Mechanisms

Rank consistent ordinal regression for neural networks with application to age estimation

Rank-consistent Ordinal Regression for Neural Networks

Fast Context Adaptation via Meta-Learning

Meta-Learning: A Survey

SynRhythm: Learning a Deep Heart Rate Estimator from General to Specific

Meta-Learning with Latent Embedding Optimization

Learning to Propagate Labels: Transductive Propagation Network for Few-Shot Learning

DeepPhys: Video-Based Physiological Measurement Using Convolutional Attention Networks

The OBF Database: A Large Face Video Database for Remote Physiological Signal Measurement and Atrial Fibrillation Detection

Dynamic Few-Shot Visual Learning Without Forgetting

On First-Order Meta-Learning Algorithms

CNN-based small object detection and visualization with feature activation mapping

Unsupervised skin tissue segmentation for remote photoplethysmography

A Simple Neural Attentive Meta-Learner

Algorithmic Principles of Remote PPG

Enhancing The Reliability of Out-of-distribution Image Detection in Neural Networks

Robust heart rate from fitness videos

Prototypical Networks for Few-shot Learning

Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks

Meta Networks

Optimization as a Model for Few-Shot Learning

Skin inhomogeneity as a source of error in remote PPG-imaging.

Decoupled Neural Interfaces using Synthetic Gradients

Self-Adaptive Matrix Completion for Heart Rate Estimation from Face Videos under Realistic Conditions

Meta-Learning with Memory-Augmented Neural Networks

Matching Networks for One Shot Learning

Ordinal Regression with Multiple Output CNN for Age Estimation

Stacked Hourglass Networks for Human Pose Estimation

Predicting Progression of Alzheimer’s Disease Using Ordinal Regression

One millisecond face alignment with an ensemble of regression trees

Remote Heart Rate Measurement from Face Videos under Realistic Situations

Microflotronic Arterial Tonometry for Continuous Wearable Non-Invasive Hemodynamic Monitoring

Detecting Pulse from Head Motions in Video

Robust Pulse Rate From Chrominance-Based rPPG

The Advantages of Wearable Green Reflected Photoplethysmography

Heart rate monitoring via remote photoplethysmography with motion artifacts reduction.

Remote plethysmographic imaging using ambient light.

Molecular prediction of disease risk and severity in a large Dutch Crohn’s disease cohort

Progressive vascular calcification over 2 years is associated with arterial stiffening and increased mortality in patients with stages 4 and 5 chronic kidney disease.

Heart rate measurement based on a time-lapse image.

Text message advertising: Response rates and branding effects

Histograms of oriented gradients for human detection

Long Short-Term Memory

Visual Heart Rate Estimation with Convolutional Neural Network

Siamese Neural Networks for One-Shot Image Recognition

A Multimodal Database for Affect Recognition and Implicit Tagging

Implicit Feedback Recommendation via Implicit-to-Explicit Ordinal Logistic Regression Mapping

Advancements in Noncontact, Multiparameter Physiological Measurements Using a Webcam

The OpenCV Library

Lack of Relationship Between Leukoaraiosis and Carotid Artery Disease

the formulation of rPPG estimation as an ordinal regression problem to cope with the mismatch in temporal domain of both visual input and collected PPG data

Non-contact, automated cardiac pulse measurements using video imaging and blind source separation

Results of average standard deviation HR measurement error on MAHNOB-HCI using different adaptation steps, L. Method SD of HR (bpm)

End-to-end (baseline)

we introduce a transductive inference method that makes use of unlabeled data for fast adaptation during deployment using a synthetic gradient generator and a novel prototypical distance minimizer

Remote Heart Rate Estimation Using a Transductive Meta-Learner 17

Field of Study

Venue Information

Name

Type

URL

Alternate Names