Simulating Mechanisms of Emotion Associated with Visual Perception of Urban Space: An Artificial Agent’s Perspective

As a collaboration between Architecture, Computer Science and Neuroscience, this paper discusses why and how the visual characteristics of urban space have an impact on human emotions, and explores how some of the underlying neural mechanisms could be simulated. The software system is tuned using machine learning so that the artificial agent’s emotional responses appear to be genuine. The system can help architects predict how humans (emotionally) experience new architectural products before they are actually realized.