Development of Creative AI for Subconscious Actions

Abstract

This dissertation involves the technical development of the Ascended Intelligence project, an AI-driven installation. The project explores real-time voicebased emotional interaction using AI, blending sound, storytelling, and symbolic feedback. The system is designed to analyze a participant’s voice tone in real time, classify their emotional frequency, and display dynamic visuals. The overall goal is to offer an emotionally intelligent experience. Human-computer interaction and artificial intelligence bring emotion recognition into the spotlight with increasing importance in developing emotions systems. Voice is a high carrier of affect information and an implicit and non-intrusive method for detecting emotions. This dissertation discusses designing and implementing an emotional state recognition system from speech in real time with openSMILE for feature extraction and classification based on machine learning