Paper Title
THERMALSENSE: AN AI-DRIVEN HUMAN TEMPERATURE DETECTION AND MONITORING SYSTEM

Abstract
Contactless temperature measurement has become a critical requirement in high-footfall environments such as airports, hospitals, and public transit hubs. Thermal Sense addresses this need through a tightly integrated pipeline that combines infrared frame capture, deep-learning-based face localization, region-of-interest thermal extraction, and time-domain signal smoothing. The system is constructed around a privacy-first philosophy, storing no biometric imagery and retaining only anonymized positional metadata alongside computed temperature readings. A field-deployable calibration routine compensates for sensor drift and emissivity variation, and a configurable alerting engine flags sustained thermal anomalies with minimal false-positive overhead. Experimental trials conducted indoors confirmed reliable face localization across typical frontal and partially occluded viewing angles, with surface temperature estimation error remaining within the tolerance band of the target sensor class. The modular codebase supports straightforward hardware substitution and site-specific threshold tuning, making Thermal Sense a viable starting point for responsible, reproducible screening deployments. Keywords - Thermal Screening, Infrared Imaging, Face Detection, Edge Computing, Temperature Estimation, Privacy-Preserving Systems, Contactless Monitoring, Signal Smoothing