Learning to Assess Heartbeat Observability for mmWave Heart-Rate Sensing

2026-10-02

AI Summary

This study addresses the signal unobservability problem in millimeter-wave radar-based heart rate sensing caused by coherent superposition of scatterers, and proposes the HEAR model. By constructing a multi-scatterer simulator to generate labels, we design a compact dual-task Transformer that jointly predicts observability and heart rate, achieving for the first time direct observability assessment from measurement data. Furthermore, a score-based selective prediction mechanism is introduced to facilitate zero-shot transfer and edge deployment. Evaluated on a real-world 120 GHz dataset, the proposed method reduces the mean absolute error to 1.6 BPM at a 50% coverage rate, with an end-to-end latency of merely 50.8 ms.

Abstract

Contactless heart-rate sensing with millimeter-wave (mmWave) radar requires assessing whether individual measurements support reliable estimation. We study learning to assess heartbeat observability, defined as the readability of the heartbeat component in an acquired phase spectrum, for selective heart-rate estimation. Coherent superposition of scatterer returns can suppress this component even under similar macroscopic observation geometry, motivating assessment directly from acquired measurements. To obtain training supervision across different observability conditions, we develop a controllable multi-scatterer frequency-modulated continuous-wave (FMCW) simulator. Agreement between the dominant heartbeat-band peak and the known heart rate provides an automatic observability label for each simulated measurement. We propose HEAR (Heartbeat Estimation with Assessed Reliability), a compact dual-task Transformer that jointly predicts an observability score and heart rate. Its input combines spectral magnitudes with frequencies relative to the respiration fundamental, providing context for respiratory harmonics. Trained solely on simulated observations, HEAR transfers zero-shot to two public real-world datasets collected at 60 and 120 GHz from 134 subjects. The same learned score supports selective prediction with both HEAR's own heart-rate head and multiple existing estimators. On the 120 GHz dataset, score-based selection reduces the HR head's mean absolute error from 17.9 BPM at full coverage to 1.6 BPM at 50% coverage. The complete pipeline achieves an end-to-end processing latency of 50.8 ms on an edge device. Project page: https://yuxuanhu9.github.io/HEAR/.

Problem

• mmWave radar
• heart-rate sensing
• heartbeat observability
• contactless sensing
• selective estimation

Innovation

• mmWave radar
• heartbeat observability
• dual-task Transformer
• zero-shot transfer
• FMCW simulator