AI for optimizing indoor visual comfort from facial analysis

Local visual discomfort in offices due to low illuminance at working stations, and direct sunlight striking computer screens or occupants’ eyes are very annoying, yet, common inconveniences. They depend on various parameters such as the orientation of the occupants, the placement of the furniture, the setup of the lighting system, the position of the windows, the location of the sun in the sky or the cloud cover. It is therefore difficult to accurately assess what local visual comfort is perceived by occupants based on fixed illuminance sensors. Consequently, artificial lighting and shading devices controlled according to these sensors can lead to poor indoor visual comfort.
This Bridging project aims at developing a face analysis system that uses machine learning to directly evaluate the subjective local visual comfort of occupants based on the video footage of their facial and eyes expressions. The face of the occupant is directly used as a visual comfort sensor. This video-based face analysis algorithm will be able to detect local visual discomfort and thus send adequate feedback for the optimum regulation of the lighting system and the shading device in an office.

https://www.ai.aau.dk/collaboration/AI+for+optimizing+indoor+visual+comfort+from+facial+analysis/