An Open Framework for Remote-PPG Methods and Their Assessment

• Published in: IEEE access Vol. 8; pp. 216083 - 216103
• Main Authors: Boccignone, Giuseppe, Conte, Donatello, Cuculo, Vittorio, D'Amelio, Alessandro, Grossi, Giuliano, Lanzarotti, Raffaella
• Format: Journal Article
• Language: English
• Published: Piscataway IEEE 2020; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Algorithms; Benchmark testing; Cognitive science; Computer Science; Computer Vision and Pattern Recognition; Datasets; Estimation; Heart rate; Heart rate variability; Image Processing; non-parametric statistical test; Nonparametric statistics; Photoplethysmography; Pipelines; Principal component analysis; pulse rate estimation; Python package; Remote photoplethysmography (rPPG); Signal and Image Processing; Skin; Sound; Statistical analysis; Statistical methods; Statistical analysis; Python package; nonparametric statistical test; pulse rate estimation; Remote photoplethysmography (rPPG)
• ISSN: 2169-3536; 2169-3536
• DOI: 10.1109/ACCESS.2020.3040936

This paper presents a comprehensive framework for studying methods of pulse rate estimation relying on remote photoplethysmography (rPPG). There has been a remarkable development of rPPG techniques in recent years, and the publication of several surveys too, yet a sound assessment of their performance has been overlooked at best, whether not undeveloped. The methodological rationale behind the framework we propose is that in order to study, develop and compare new rPPG methods in a principled and reproducible way, the following conditions should be met: 1) a structured pipeline to monitor rPPG algorithms' input, output, and main control parameters; 2) the availability and the use of multiple datasets; and 3) a sound statistical assessment of methods' performance. The proposed framework is instantiated in the form of a Python package named pyVHR (short for Python tool for Virtual Heart Rate), which is made freely available on GitHub ( github.com/phuselab/pyVHR ). Here, to substantiate our approach, we evaluate eight well-known rPPG methods, through extensive experiments across five public video datasets, and subsequent nonparametric statistical analysis. Surprisingly, performances achieved by the four best methods, namely POS, CHROM, PCA and SSR, are not significantly different from a statistical standpoint higighting the importance of evaluate the different approaches with a statistical assessment.