Інтелектуальне подання мультимедійних даних шляхом нейромережевої трансформації візуального стилю

Abstract

This study explores the application of artificial intelligence for the dynamic stylization of multimedia interfaces, focusing on the visual code of the game Persona 5. The research identifies the key semantic elements of high-contrast UI design and proposes a real-time transformation framework based on Fast Style Transfer architecture. The study analyzes factors influencing performance, such as GPU latency and content readability, providing a method for emotional and adaptive data presentation in modern multimedia systems. The implementation results demonstrate that combining modern neural network architectures with gaming design technologies provides high user engagement and a unique interaction experience