Can computers visually quantify human thermal comfort?: Short Paper

摘要: The main driving factor in determining control settings of an air conditioning system is indoor thermal condition and its association with occupants' satisfaction. In recent years, the ubiquity computing devices has led to techniques for learning personalized comfort by using feedback through smartphones or (commonly wearable) sensors measure physiological variables. this paper, seeking a non-contact non-intrusive approach, we are investigating whether computers could visually identify human digital videos. This investigation been conducted form developing specialized image processing algorithmic framework, which relies on Eulerian magnification approach. approach amplifies subtle changes signals that associated thermoregulation. proposed framework evaluated experimental studies subjects chamber hot cold conditions. quantified results demonstrated effectively identified thermoregulation states subjects.