Creativity-Model | Takato Horii | Osaka University

Co-Creative Music Generation through Interactive Preference Estimation

Mon, 01 Jan 2024 00:00:00 +0000

We study co-creative music generation that estimates users’ affective preferences on the fly through dialogue and reflects them in diverse creative suggestions. Using In-Context Learning and In-Context Preference Learning, we take in value criteria without additional training and aim to support humans and AI in collaboratively creating new musical expressions.

Creative Data Generation Using Generative Adversarial Networks

Mon, 01 Jan 2024 00:00:00 +0000

Using generative adversarial networks (GANs), we propose a framework for creative data generation that goes beyond mere reproduction of training data. Through the adversarial training of generator and discriminator, we build mechanisms that acquire novel samples and aim at a computational understanding of creative behavior in robots and artificial agents.

Multi-Agent Simulation of the Systems Model of Creativity

Mon, 01 Jan 2024 00:00:00 +0000

We implement Csikszentmihalyi’s systems model of creativity (the interaction of individual, field, and domain) as an in silico simulation by multiple generative-AI-driven agents, and constructively clarify how creativity emerges on a social system. We further address the co-evolution of generative strategies and evaluative preferences in AI music (cultural drift), aiming at a multifaceted understanding of creative dynamics in agent populations.