[B-19] Healthcare and Medical Information Communication Technology

10:30 AM - 10:45 AM JST(1:30 AM - 1:45 AM UTC)Candidate for Academic Incentive AwardCareer Explorer

[B-19-04] Non-Contact Heart Rate Estimation in Low-SNR Environments Using a
Low-Power 60 GHz MIMO-FMCW Radar Based on CNN Deep Learning: A Study
on Training Data Optimization and Distance Dependency

◎△Rinto Ogawa¹, Yudai Imada², Takeshi Toda¹ (1. College of Science & Technology, NIHON UNIVERSITY, 2. Graduate School of Science & Technology, NIHON UNIVERSITY)

Keywords:

Heart Rate,Vital Signs,Healthcare,Millimeter-Wave Sensor,Sensor

In our previous work, we investigated whether a 1D ResNet model could achieve high-accuracy heart rate estimation even in environments with a low signal-to-noise ratio. At a distance of 1 m, the RMSE values for the low, medium, and high heart rate ranges were 9.87, 6.72–7.70, and 16.77 bpm, respectively. In this paper, we report an evaluation conducted by revising the training data (subject data) and extending the measurement distance up to 3 m.