Google has unveiled a research system that uses a smartphone’s front-facing camera and artificial intelligence to estimate heart rate and resting heart rate during normal device use. Called Passive Heart Rate Monitoring During Smartphone Use in Everyday Life (PHRM), the technology captures short facial videos after a user unlocks their phone and analyzes subtle changes in light reflected from the skin as blood flows through facial blood vessels. By combining camera-based sensing with AI models, the system aims to turn everyday smartphone interactions into opportunities for passive health monitoring without requiring additional devices.
The platform relies on photoplethysmography, a technique commonly used in wearable health devices, and processes eight-second video clips through lightweight AI models running directly on the smartphone. These models generate heart rate estimates and confidence scores, which are aggregated throughout the day to calculate resting heart rate. Google trained and evaluated the system using more than 350,000 video samples collected from nearly 700 participants across both controlled and real-world settings, with efforts made to ensure performance across diverse skin tones.
According to the study published in Nature, the technology achieved heart rate measurements with less than 10% average error compared to electrocardiogram readings and produced resting heart rate estimates comparable to consumer fitness wearables. Google says future development will focus on improving video stabilization, camera exposure and motion detection to enhance accuracy and capture measurements during optimal resting moments. The research highlights the growing potential for smartphones to serve as passive health monitoring tools without the need for dedicated sensors or wearable devices.
Click here to read the original news story.
