Monitoring Sleep Posture Based on Temperature and Humidity Sensor
- Resource Type
- Conference
- Authors
- Kim, Hyung-Ju; Jun, Won-Ho; Shin, Woo-Chul; Hong, Youn-Sik
- Source
- 2024 International Conference on Artificial Intelligence in Information and Communication (ICAIIC) Artificial Intelligence in Information and Communication (ICAIIC), 2024 International Conference on. :022-025 Feb, 2024
- Subject
- Communication, Networking and Broadcast Technologies
Components, Circuits, Devices and Systems
Computing and Processing
Fields, Waves and Electromagnetics
Power, Energy and Industry Applications
Robotics and Control Systems
Signal Processing and Analysis
Temperature sensors
Temperature measurement
Temperature distribution
Sleep
Humidity measurement
Humidity
Machine learning
sleep posture
temperature and humidity sensor
CNN
data labeling
- Language
- ISSN
- 2831-6983
In this paper, we proposed a method of attaching seven temperature and humidity sensors to a pillow and monitoring changes in temperature and humidity to determine sleeping posture. The existing method of determining sleep position is to monitor changes in pressure using FSR sensors mounted inside a pillow. Changes in pressure are limited to simply identifying changes in head movement, but changes in temperature and humidity have the advantage of being able to identify physical changes in the subject. For machine learning modeling, the temperature and humidity dataset accumulated during sleep was subjected to a data labeling process through discrimination of sleep posture images. With the CNN-based modeling the accuracy of discriminating sleep posture reached about 95%. This confirms that temperature and humidity sensor-based sleeping posture monitoring is sufficiently applicable.