Skip to main content

Overview

When a Muse headband with PPG is available alongside camera-based rPPG, the SDK supports sensor fusion, using the contact PPG as a reference to calibrate and improve camera-based heart rate estimates.

MuseFusionCalibrator

Fuses Muse PPG readings with camera rPPG estimates for improved accuracy:
The RppgProcessor also has a built-in Muse integration:

MuseCalibrationModel

A simple regression model that learns the relationship between camera and contact BPM over time:

BPM Evidence Types

The processor produces evidence from multiple analysis methods:

Muse PPG Filter

Apply Muse-style bandpass filtering to PPG samples:

Calibration Workflow

  1. Start with camera-only rPPG
  2. When a Muse headband connects, feed its PPG as ground truth
  3. The calibration model learns the camera-to-contact mapping
  4. Once trained, camera-only estimates are corrected using the learned model
  5. The fusion calibrator combines both sources for maximum accuracy

Next

Frame Sources

Camera capture and face detection

rPPG Camera Integration

End-to-end webcam pipeline