Contrast-Phys+: Unsupervised and Weakly-supervised Video-based Remote Physiological Measurement via Spatiotemporal Contrast
This addresses the challenge of obtaining ground truth signals for remote health monitoring, offering a more practical solution with incremental improvements in efficiency and robustness.
The paper tackles the problem of video-based remote physiological measurement (rPPG) by proposing Contrast-Phys+, a method that can be trained unsupervised or weakly-supervised to avoid costly ground truth data, and it outperforms state-of-the-art supervised methods on five datasets.
Video-based remote physiological measurement utilizes facial videos to measure the blood volume change signal, which is also called remote photoplethysmography (rPPG). Supervised methods for rPPG measurements have been shown to achieve good performance. However, the drawback of these methods is that they require facial videos with ground truth (GT) physiological signals, which are often costly and difficult to obtain. In this paper, we propose Contrast-Phys+, a method that can be trained in both unsupervised and weakly-supervised settings. We employ a 3DCNN model to generate multiple spatiotemporal rPPG signals and incorporate prior knowledge of rPPG into a contrastive loss function. We further incorporate the GT signals into contrastive learning to adapt to partial or misaligned labels. The contrastive loss encourages rPPG/GT signals from the same video to be grouped together, while pushing those from different videos apart. We evaluate our methods on five publicly available datasets that include both RGB and Near-infrared videos. Contrast-Phys+ outperforms the state-of-the-art supervised methods, even when using partially available or misaligned GT signals, or no labels at all. Additionally, we highlight the advantages of our methods in terms of computational efficiency, noise robustness, and generalization. Our code is available at https://github.com/zhaodongsun/contrast-phys.