NOVEL TO ANIMATION: A LIGHT NOVEL BASED PHOTOGRAPHIC PROJECT

Abstract

The development of innovative technologies in a variety of industries has been greatly aided by artificial intelligence (AI). The anime industry is one such area where AI- powered solutions can be applied to produce animation in an organized and effective manner. In this research, we build a system that utilize the Stable Diffusion AI model to create anime sequences with relevant descriptions scraped from light novels. Arguably, anime artists face the worst of it given the weekly anime release cycle with its associated stress, sleep deprivation, and health problems. Our approach seeks to enable their work with greater precision and efficiency through AI-powered image generation. The system comprises four constituent modules: database, input processing, image creation, and user interface. The processing module removes unwanted characters from the text description of the scene to enhance processing efficiency. The picture generation module uses the Stable Diffusion AI model text-to-image mapping to create an image depicting a latent vector representation of the intended scene. The user interface module allows users to enter text descriptions and view the generated anime scenes from their own perspective. Compared to existing technologies, our suggested approach enables greater accuracy and efficiency. Input scene descriptions from the related light novel into our system and the text will be converted into realistic and accurate anime images within a short time using Stable Diffusion AI model, making the process more efficient.