WISSH – Wallpaper Innovator by Synthesizing Song Hallmark

This paper presents an intelligent system that transformsaudio music into visually expressive animated wallpapers.Leveraging deep learning-based mood recognition and advancedimage generation via Stable Diffusion, the project combinesacoustic feature extraction with a culturally aware mood classifierto interpret musical emotions. These emotions are converted intostylistic prompts that guide the creation of aesthetic, animatedwallpapers. The system integrates PyTorch for mood detection,HuggingFace Transformers for audio embeddings, and Diffusersfor visual generation. The platform offers both single and batchprocessing, enabling immersive visual experiences from soundinput. By bridging the gap between auditory emotion and visualstorytelling, this project opens a pathway for more immersivemultimedia experiences. The paper explores both the algorithmicdesign and user-centric features of the system, demonstratingpotential applications in art therapy,music streaming platforms, and personalized device aesthetics.