Robust and Generalizable Heart Rate Estimation via Deep Learning for Remote Photoplethysmography in Complex Scenarios
This work addresses the need for reliable, non-contact heart rate monitoring in healthcare and fitness applications, but it is incremental as it builds on existing deep learning methods with specific improvements.
The paper tackled the problem of inaccurate and non-robust heart rate estimation from facial videos in complex scenarios using remote photoplethysmography, achieving a mean absolute error of 7.58 on a challenging dataset, which outperformed state-of-the-art models.
Non-contact remote photoplethysmography (rPPG) technology enables heart rate measurement from facial videos. However, existing network models still face challenges in accu racy, robustness, and generalization capability under complex scenarios. This paper proposes an end-to-end rPPG extraction network that employs 3D convolutional neural networks to reconstruct accurate rPPG signals from raw facial videos. We introduce a differential frame fusion module that integrates differential frames with original frames, enabling frame-level representations to capture blood volume pulse (BVP) variations. Additionally, we incorporate Temporal Shift Module (TSM) with self-attention mechanisms, which effectively enhance rPPG features with minimal computational overhead. Furthermore, we propose a novel dynamic hybrid loss function that provides stronger supervision for the network, effectively mitigating over fitting. Comprehensive experiments were conducted on not only the PURE and UBFC-rPPG datasets but also the challenging MMPD dataset under complex scenarios, involving both intra dataset and cross-dataset evaluations, which demonstrate the superior robustness and generalization capability of our network. Specifically, after training on PURE, our model achieved a mean absolute error (MAE) of 7.58 on the MMPD test set, outperforming the state-of-the-art models.