Remote photoplethysmography (rPPG) in the wild: Remote heart rate imaging via online webcams.

Abstract

Remote photoplethysmography is a low-cost technique to measure physiological parameters such as the heartrate by analysing videos of a person. There has been growing attention in this technique due to the increased possibilities and demand for running psychological experiments in online platforms. Many progresses have been made due to technological advancements in commercially available cameras and in the algorithms used to process the videos. Despite these advancements, past research has shown that suboptimal conditions of the video recordings can severely impair the accuracy of the rPPG. Our aim was to test the accuracy of a modified open-source rPPG algorithm on videos collected via an online platform, without control on the hardware of the participants and the contextual variables, such as illumination, distance and motion. Across two experiments, we compared the results of the rPPG extraction algorithm to a validated dataset used for rPPG testing, and then we collected 231 online video recordings and compared the results of the rPPG extraction against a validated mobile heart rate application. Results showed that the rPPG was highly accurate on both datasets thus providing an improved tool for recording and analyzing heartrate in online experiments.