Research on Moving Webtoons Using Generative AI

Following the rise of multi-modal generative models, many industries, including webtoon production, have begun leveraging artificial intelligence (AI) tools for efficient content creation. In this paper, we present a novel framework aimed at improving webtoon creation through the integration of generative AI technologies. Our approach targets two critical challenges: maintaining consistency in character design and improving Korean language support in generative models. To address these challenges, we propose the development of a customized Stable Diffusion model that supports Korean language inputs, alongside an AI system optimized for webtoon production workflows. This framework ensures consistent character designs while expanding the accessibility of AI tools for Korean creators. Furthermore, we introduce techniques for incorporating dynamic motion effects into static webtoon imagery, resulting in an immersive content experience. By establishing an AI- powered production pipeline, our proposed framework reduces the production time while preserving the creative quality. The proposed framework has the potential to elevate both the efficiency and competitiveness of the webtoon industry in the era of AI-driven content creation.