StreamPPG: Low-Latency rPPG Estimation via Consistent Privileged Learning

2026-06-22

AI Summary

This work addresses the limitations of existing frame-wise rPPG methods, which struggle to model the long-term periodicity of physiological signals and thus suffer from suboptimal accuracy, while clip-based approaches incur high latency. To reconcile low latency with high accuracy, we propose StreamPPG—a lightweight streaming architecture that leverages a Consistent Privileged Learning (CPL) strategy during training, where ground-truth rPPG signals serve as privileged information to enhance feature representation. Our method achieves state-of-the-art performance across multiple benchmark datasets and enables real-time inference on edge devices, matching the accuracy of clip-level methods while preserving the low-latency advantage of frame-wise processing.

Abstract

Remote photoplethysmography (rPPG) estimates the blood volume pulse (BVP) signal from facial videos, enabling contact-free health monitoring. Conventional clip-wise approaches, which use video clips as input, require capturing over one hundred frames before inference, thus introducing several seconds of delay and hindering real-time use. Meanwhile, frame-wise approaches struggle to capture long-range temporal and periodic features of physiological rhythms, and therefore lead to reduced estimation accuracy. To overcome these issues, we propose StreamPPG, a unified architecture that enables low-latency frame-wise physiological signal estimation while achieving competitive accuracy compared with clip-wise approaches. StreamPPG is trained under a consistent privileged learning (CPL) strategy, which leverages ground-truth rPPG signals as privileged information to enhance the model's representation capability. Extensive experiments demonstrate that StreamPPG achieves state-of-the-art accuracy across multiple datasets while maintaining real-time throughput on edge devices.

Problem

• remote photoplethysmography
• low-latency
• real-time estimation
• temporal features
• physiological signal

Innovation

• StreamPPG
• rPPG
• privileged learning
• low-latency
• real-time physiological monitoring