Abstract
What makes a design beautiful? Design styles developed during past eras such as byzantine, classical, gothic, renaissance and baroque are universally admired as being cultural icons and are widely appreciated by people from all walks of life. Throughout the years, many philosophers, architects and physicists have come up with theories and frameworks to measure the subjective topic of aesthetics, but none stood out like Birkhoff's aesthetic measure which used a mathematical approach to quantifying beauty. In this paper, we investigate the aesthetic appeal of generative AI model outputs, trained on datasets recognized for their aesthetic quality, and by employing biometric data analysis to cross-reference these results with Birkhoff's aesthetic measurement framework. Stemming from Neuroarchitecture, wearable technologies offer an insight into the correlation between spatial qualities and human perception that can be extended into aiding us, architects in designing better for the built environment. In our experiment, we generated a set of interior images in assorted styles following current interior design trends. The generated outputs are first scored based on Birkhoff's measurements of aesthetics and cross referenced with data obtained from wearable technologies such as an eye tracker and electroencephalogram (EEG) headset. Eye tracking glasses can detect fixations, saccade patterns, and pupil dilation, which can reflect subconscious thoughts from the user. The EEG is also utilised to complement the eye tracking data as a means to reflect on positive or negative impressions towards a particular subject. Overall, this innovative approach adapts Birkhoff's aesthetic measurement in a human-centric and evidence-based way, providing architects with a framework to systematically evaluate design. It merges Birkhoff's theorem with unbiased subconscious metrics to compare current and historical aesthetic trends, and behavioural research to pinpoint common aesthetic preferences. This method also leverages biometric data to align architectural design more closely with user perspectives, breaking down traditional communication barriers and offering clearer insights into client preferences.