HCLSensor: Human Comfort Level Sensor

Human comfort level sensing, crucial for applications in healthcare and smart home applications, often relies on wired or wireless sensors. However, using these devices poses challenges such as non-subjective feelings, maintenance costs, energy consumption, and negative environmental impact. To tackle these challenges, we for the first time propose a novel approach - Human Comfort Level Sensor. In this system, we leverage existing wireless communication infrastructure to discern the user's comfort level. The key innovation lies in the fact that wireless signals can capture subtle reactions in the human body that indicate changes in comfort level. By exploring the relationship between wireless signals and human comfort level, our HCLSensor system is designed to gather wireless signals, extract distinctive features, and use a deep learning model to detect human comfort level. More importantly, the HCLSensor can be trained with minimal data from one location, such as a lab, yet applied in another, like a residential home, or trained on one individual's data and used with another's. Real-world experiments demonstrate that the HCLSensor system achieves significant accuracy: up to 96% when training and testing occur at the same location, 94% when in different places, and 84% when utilizing data from different individuals.