Heart Rate Estimation Algorithm for Radar Signals Based on Resnet18 and MobileViT

Non-invasive heart rate monitoring technology using millimeter wave radar offers high comfort, high convenience, and wide applicability to different user groups. However, current heart rate monitoring algorithms are likely to be affected by environmental noise, DC offset, and respiratory harmonics, which can compromise result accuracy. To address these issues, this paper proposes a radar signal heart rate estimation algorithm based on Resnet18 and MobileViT. In the phase preprocessing stage, techniques such as phase difference calculation, interpolation smoothing, and wavelet threshold denoising are employed on the phase signal. During the heart rate estimation stage, the Resnet18 and MobileViT networks are utilized to process the phase signal. Experimental analysis indicate improved heart rate estimation accuracy with the proposed method..