Lecture
This lecture discusses the challenges of monitoring occupants' thermal comfort in buildings. The instructor highlights the limitations of traditional methods, such as environmental sensors and surveys, which often fail to capture individual experiences accurately. The emergence of wearable devices is noted, but their intrusive nature raises privacy concerns. To address these issues, the lecture introduces a framework developed in the ISCOM project that utilizes computer vision and machine learning to monitor thermal states remotely. By employing a network of cameras, the framework predicts local skin temperatures and core body temperatures without storing personal data. The instructor explains the use of various computer vision algorithms to detect features such as age, gender, and clothing, which are essential for accurate thermal comfort assessments. The lecture concludes with a summary of the project's accuracy in predicting personal features and temperatures, emphasizing the interdisciplinary collaboration between two laboratories to create a scalable and non-intrusive monitoring solution.