Developing an Intermediate Framework for Enhancing Comic Creation Through Generative AI

This paper endeavors to explore the challenges and opportunities arising in the collaborative creation of comics involving cartoonists and Generative Artificial Intelligence (GAI), specifically the AI Image Generator, such as stable diffusion. Presently, AI Image Generators heavily rely on text prompts and reference images. However, these elements lack direct alignment with the story scripts and sketches of cartoonists. Consequently, cartoonists are required to gain a substantial understanding of AI generation, leading to extensive communication between cartoonists and AI engineers. To comprehensively understand and analyze the current role of AI in comic creation, we conducted a focus group discussion involving both cartoonists and AI engineers. Subsequently, we provide detailed insights into the design of an upcoming comic generation workshop scheduled for the next two weeks. Expanding upon the aforementioned research, we suggest design elements for the intermediate layer that bridges comic creation and AI generation. To validate these design elements, the paper utilizes the AI comic “MBTI: i and e personalities” as a case study. Our conclusion underscores that establishing an intermediate layer between comic creation and AI generation, coupled with the effective design of corresponding elements, can significantly enhance the control and precision of AI comic image generation.