COVID-19 drastically raised the importance of non-contact based healthcare methods. Low blood oxygen levels of a person, which can be unnoticeable, are potentially a precursor of COVID-19. Contact based methods for measuring blood oxygen saturation could spread the contagious disease. Therefore, this paper investigates non-contact RGB camera-based peripheral oxygen saturation estimation by remote photoplethysmography (rPPG) methods. The novel aspects of non-contact oxygen saturation that we are looking into are: (1) Applying SpO$_2$ predictor neural networks to rPPG signals obtained from facial regions, instead of the less practical hand based skin regions. To be more specific, we show in a facial based setting that in the relatively uncontrolled environment the traditional Ratio-of-Ratios pulse oximetry principles fail. In the leave-one-participant-out experiments, the RoR method achieved a correlation of $-0.05$, whereas neural networks showed the capability of dealing with the inherent challenges of the PURE dataset by achieving a superior correlation of $0.64$. These challenges are lighting variation due to subtle head motion and clouds alternatively blocking the sun. (2) The first end-to-end neural networks for SpO$_2$ estimation are introduced by replacing traditional hard pixel region-of-interest selectors, which assign equal weight to each selected pixel, with convolutional soft-attention masks. (3) By using an adapted version of a recent heart and breathing rate estimator network, called DeepPhys, we indicate that the current state-of-the-art is far from optimal. This is done by comparing the window-based constructed end-to-end neural networks with Adapted DeepPhys, which is based on single frame differences. Finally, our research\footnote{Code available on \url{https://github.com/jimkok9/oxygenSaturation}} shows that non-contact facial based SpO$_2$ estimation by RGB camera remains a difficult task. However, as our results indicate, more sophisticated deep learning model might become a viable diagnostic tool for this task in the future.

Contactless measurement of changes in blood volume by exploiting the color fluctuations in the face is a technique commonly referred to as remote photoplethysmography (rPPG). Recent developments show promising results for heart rate estimation from low-cost cameras, making applications in remote healthcare possible. Remote PPG applications in at home diagnostics focus predominantly on heart rate estimation, while other features of the blood volume pulse can provide valuable information as well, but have scarcely been studied using rPPG. In this work, we aim to lay a foundation for rPPG feature prediction. We study pulse wave prediction using a variety of input representations, model architectures and datasets to thereby investigate which combined approaches are the most promising. Our results show which input representation is most suitable based on the feature of interest and demonstrate the ability to predict pulse waveform features using rPPG. These results take the first steps towards including the prediction of a wide range of waveform properties to make more remote health monitoring possible.