Indoor human counting based on range-Doppler spectrum in the presence of material penetration and occlusion

This paper introduces an ensemble approach for indoor human counting method based on millimeter-wave (mmWave) radar technology. Through advanced spatio-temporal range-Doppler analysis, the proposed model classifies room occupancy (empty, single-person, multiple people) with 90% accuracy. Beyond precision counting, the system's real-time data facilitates smart applications, such as optimizing air conditioning during pandemics or other dynamic scenarios. The most common technologies such as image or wearable inertial sensors are not preferable due to privacy issues or subject discomfort. This compact solution combines mmWave radar and deep learning for efficient, adaptable, and secure indoor environment management.