본문 바로가기
  • Home

Hybrid Heart Rate Measurement Combining Deep Learning PPG Classification and Peak Detection

  • Journal of The Korea Society of Computer and Information
  • Abbr : JKSCI
  • 2025, 30(11), pp.269~275
  • Publisher : The Korean Society Of Computer And Information
  • Research Area : Engineering > Computer Science
  • Received : September 23, 2025
  • Accepted : October 31, 2025
  • Published : November 28, 2025

Sejong Lee 1 Young-Bok Cho 2

1영남대학교
2국립경국대학교

Accredited

ABSTRACT

The recent popularization of wearable devices such as smartwatches has enabled users to monitor various biometric data, such as heart rate and oxygen saturation, in real time. However, existing commercial devices have limitations such as low measurement accuracy, errors occurring under certain conditions, and difficulty in clearly identifying error areas. Therefore, in this study, we developed a highly accurate biometric signal measurement device based on a PPG sensor to improve these issues. Deep learning was used to classify PPG signals, and a signal processing algorithm was applied to calculate heart rate. Experimental results showed that the deep learning prediction accuracy was 97.46%, and the heart rate measurements showed similar trends to those of commercial wearable devices, confirming high reliability.

Citation status

* References for papers published after 2024 are currently being built.