What does this research mean for the field?

The CHROM algorithm consistently yields the lowest error and highest stability for heart-rate estimation from RGB datasets, particularly when paired with tree-based models. Novelty: ClaimNovelty.SYNTHESIS. Consensus alignment: ConsensusAlignment.NEUTRAL.

What question did this study set out to answer?

The research aims to create a unified framework for evaluating heart rate estimation methods using remote photoplethysmography.

January 10, 2026Electronics0 citationsOpen Access

A Unified Benchmarking Framework for Classical Machine Learning Based Heart Rate Estimation from RGB and NIR rPPG

SQSahar QaadanGerman Jordanian University GJGhassan Al Al JayyousiGerman Jordanian University AAAdam AlkhalailehGerman Jordanian University

Key Result

The CHROM algorithm consistently yielded the lowest error and highest stability for heart-rate estimation from RGB datasets, particularly with tree-based models.

Key Points

The research aims to create a unified framework for evaluating heart rate estimation methods using remote photoplethysmography.
Standardized ROI extraction and signal preprocessing across four datasets.
Benchmarking five rPPG extraction methods combined with four classical regressors.
Utilizing MAE, RMSE, and R2 for performance evaluation.
Incorporated permutation feature importance for interpretability.
CHROM algorithm outperforms others across RGB datasets with minimal error.
PCA with spatial patch decomposition significantly outperforms ICA for NIR recordings.
Classical models provide interpretable baselines but face limitations in generalization.

Structured PICO

Population

Four publicly available datasets for remote photoplethysmography (rPPG): UBFC-rPPG Part 1, UBFC-rPPG Part 2, VicarPPG-2, and IMVIA-NIR

Intervention

Five rPPG extraction methods (Green, POS, CHROM, PBV, PCA/ICA) combined with four classical regressors

Comparator

Comparison among the different extraction methods and regressors

Outcome

Heart rate estimation accuracy measured by MAE, RMSE, and R2

This benchmarking framework establishes robust cross-dataset baselines for classical machine learning-based heart rate estimation from rPPG, identifying CHROM and PCA as optimal methods for RGB and NIR modalities, respectively.

Main Result

Absolute Event Rate: 0% vs 0%

Abstract

This work presents a unified benchmarking framework for evaluating classical machine-learning–based heart-rate estimation from remote photoplethysmography (rPPG) across both RGB and near-infrared (NIR) modalities. Despite extensive research on algorithmic rPPG methods, their relative robustness across datasets, illumination conditions, and sensor types remains inconsistently reported. To address this gap, we standardize ROI extraction, signal preprocessing, rPPG computation, handcrafted feature generation, and label formation across four publicly available datasets: UBFC-rPPG Part 1, UBFC-rPPG Part 2, VicarPPG-2, and IMVIA-NIR. We benchmark five rPPG extraction methods (Green, POS, CHROM, PBV, PCA/ICA) combined with four classical regressors using MAE, RMSE, and R2, complemented by permutation feature importance for interpretability. Results show that CHROM is consistently the most reliable algorithm across all RGB datasets, providing the lowest error and highest stability, particularly when paired with tree-based models. For NIR recordings, PCA with spatial patch decomposition substantially outperforms ICA, highlighting the importance of spatial redundancy when color cues are absent. While handcrafted features and classical regressors offer interpretable baselines, their generalization is limited by small-sample datasets and the absence of temporal modeling. The proposed pipeline establishes robust cross-dataset baselines and offers a standardized foundation for future deep-learning architectures, hybrid algorithmic–learned models, and multimodal sensor-fusion approaches in remote physiological monitoring.

Ask AI

Helpful

Bookmark

View Full Paper