The integration of artificial intelligence (AI) in creative disciplines has heralded a new era of digital artistry, facilitating the transition from rudimentary sketches to fully realized figures. Present AI drawing tools, predominantly employed as adjuncts in editing, have yet to achieve autonomy in image generation. Current methods tend to culminate in finalized art pieces without an illustrative journey from inception to completion.

In previous studies, we demonstrated the capability to generate body blocks from anime character images; however, these often failed to match the poses of the input images. In this study, we address this limitation by refining our model to more accurately align the generated body structures with the intended poses. Addressing this gap, we present a novel AI-driven framework that harnesses the power of Generative Adversarial Networks (GANs), Variational Autoencoders (VAEs), and Sequence-to-Sequence (Seq2Seq) models. Our system accepts an anime character image as input and produces a sequence of four progressive illustrations: a corresponding skeleton, body block, muscle structure, and refined line art. This staged generation mirrors the authentic process of artistic creation, offering a comprehensive educational tool for novice artists. The purpose of this research is to develop a comprehensive system that furnishes beginners with a detailed, step-by-step guidance in drawing, demystifying the intricate art of illustration.

This initiative not only advances AI in art but also provides a unique pedagogical approach that details the intermediate stages of drawing, which are often overlooked by existing applications. By enabling beginners to visualize each step, our system serves to shorten the learning curve and enrich the study of drawing. Our empirical evaluations indicate a significant potential for our method to empower aspiring artists, providing a scaffold to elevate their rudimentary skills to create complete and anatomically coherent figures.

This work was supported by Cabinet Office, Government of Japan, Cross-ministerial Moonshot Agriculture, Forestry and Fisheries Research and Development Program, "Technologies for Smart Bio-industry and Agriculture" (funding agency: Bio-oriented Technology Research Advancement Institution).