Research & Validation
The studies, benchmarks, and ongoing validation behind Circadify's contactless rPPG models.
Comprehensive Model Validation
We are actively developing and evaluating our rPPG models across the health metrics we measure. Our ongoing research involves benchmarking against reference devices under a range of real-world conditions.
Select a metric to learn more about our research approach and current findings.
Request Validation Report
Select a metric and we'll send you detailed validation data.
Our biggest advantage is our data
Camera-based vital signs are only as good as their data. We built one of the most rigorous and representative datasets in the field, and it is what makes our measurements hold up outside the lab, on the phones people actually use.
Lab accuracy that holds up in the real world
Every participant is recorded at the same moment on lab-grade reference cameras and on ordinary iPhone and Android phones. Accuracy we prove in the lab carries over to the phone in someone's hand.
Measured against clinical ground truth
Readings are checked against medical reference instruments for heart rate, heart rate variability, blood pressure, oxygen saturation, and respiration, so every data point is anchored to a real clinical measurement.
Built for every skin tone
Our data spans the full Fitzpatrick I-VI range and deliberately over-samples darker skin, the blind spot that causes most camera-based vitals to fall apart outside the lab.
Owned, consented, and defensible
The dataset is proprietary, collected with consent for commercial use, and owned or exclusively licensed by Circadify. Competitors can't simply download or replicate it.
Validation from an independent research group
Circadify's performance is being evaluated by an independent, third-party clinical research organization. The results come from outside the company, so partners and investors don't have to take our word for it.
Results are expected in 2026, with a regulatory submission to follow. Circadify supports monitoring and screening, and is not intended for diagnosis.
Request preliminary data →Published rPPG Research
Explore the foundational research behind remote photoplethysmography technology
TYrPPG: Uncomplicated and Enhanced Learning Capability rPPG for Remote Heart Rate Estimation
A novel approach to simplify rPPG models while enhancing learning capability for accurate remote heart rate estimation.
Read Paper →Gaze into the Heart: A Multi-View Video Dataset for rPPG and Health Biomarkers Estimation
Introduces a comprehensive multi-view video dataset for advancing rPPG research and health biomarker estimation.
Read Paper →Advancing Generalizable Remote Physiological Measurement through the Integration of Explicit and Implicit Prior Knowledge
Proposes methods to improve generalization in remote physiological measurement by combining explicit and implicit prior knowledge.
Read Paper →Reperio-rPPG: Relational Temporal Graph Neural Networks for Periodicity Learning in Remote Physiological Measurement
Leverages graph neural networks to capture temporal relationships for improved periodicity learning in rPPG signals.
Read Paper →Deep Learning-based Remote Photoplethysmography Methods: A Review
Comprehensive review of deep learning approaches for camera-based physiological measurement and vital sign estimation.
Read Paper →Evaluation of Biases in Remote Photoplethysmography Methods
Critical analysis of biases present in rPPG methods across different demographics and lighting conditions.
Read Paper →