An aesthetic emotion metric model for understanding the emotional reception of AI portrait art

The research attempts to resolve a recognized limitation within the field of cognitive science of art (historical and now AI-based). The current domain lacks a meaningful metric (or model) to measure/categorize the "aesthetic emotion" reception of perceived art. The main contribution of this thesis is the creation, adaptation, and validation of a new emotional model and instrument. We have designed and implemented the Visual Aesthetic Wheel of Emotion (VAWE) in order to measure 'aesthetic emotional' reactions to art. While this system has the potential to benefit all types of art, this research limits our measurements and work to portrait art. We conducted 4 studies to: create VAWE's structure and descriptors, then validate it with both historical portrait art and AI generated portrait art from our system. The research also establishes and refines a Generative AI Portraiture System capable of producing fine art quality portraits as perceived by our users. The system incorporates a modular and cognitive based generative approach to AI art making. We then tie this AI portraiture system to our VAWE system to create a prototype Emotionally Aware Portrait System. This research aims to categorize portrait art generated from our system based on aesthetic emotion and discuss the development of a system capable of generating art that elicits desired emotional responses. This can have benefits in entertainment, the arts and health. In all our methods and approaches we attempt to recognize and address ethical issues in AI, datasets, and authorship issues in AI art.