⚡ MuGen V2: Image to Music Matrix ⚡

MuGen V2 is a cyberpunk-themed web application built with Gradio that transforms images into music. The system utilizes image captioning networks, LLM prompt engineering, and bleeding-edge audio generation APIs to compose soundtrack matrices tailored to the mood of an image.

🔮 Core Features

• Visual Spectrography: Image processing using the Kosmos-2 visual language model.
• Neural Prompt Alignment: Uses zephyr-7b via an inference pipeline to interpret visual metadata into sophisticated musical compositions/chords.
• Multi-Engine Audio Synthesis: Integrated routing support for AudioLDM-2, Riffusion, Mustango, and more.
• Premium Cyberpunk UI: Custom Neon injected CSS styling for a high-end terminal experience.

🚀 Getting Started

Prerequisites

Make sure you have Python 3.10+ and a Hugging Face API Token.

Installation

1. Clone the repository:

   git clone [https://github.com/YOUR_USERNAME/MuGen-V2.git](https://github.com/YOUR_USERNAME/MuGen-V2.git)
   cd MuGen-V2
   

Other Credits

Used ffilioni API clients from HuggingFace.

Preview

Preview on Hugging Face: https://huggingface.co/spaces/eva1267890/MuGen Note: You may or may not need a HuggingFace account for viewing this space.

Screenshot