MangaNarrator 📚➡️🎬

Turn your Manga PDFs into witty, narrated "Read-Along" videos instantly.

MangaNarrator is an AI-powered tool that automatically converts manga PDF pages into dynamic video content. It uses Google Gemini 1.5 Flash to analyze panels and generate a witty script, Deepgram Aura for lifelike voice synthesis, and FFmpeg to stitch it all together into a seamless video.

🚀 Features

• AI Vision & Scripting: Uses Gemini 1.5 Flash to "see" the manga page, understand the panel flow (Right-to-Left), and generate a script with a specific persona (e.g., Witty Host, Serious Narrator).
• "Ultimate Narrator" Logic: Automatically inserts narrative hooks, facial analysis, and action descriptions between dialogue for a cinematic experience.
• High-Quality TTS: Powered by Deepgram Aura for near-human voice generation.
• Parallel Processing: Multi-threaded backend processes all pages concurrently for fast video generation.
• Neurobrutalist UI: A distinct, high-contrast frontend design.

🛠️ Tech Stack

• Backend: Python, Flask
• AI/LLM: Google Gemini API (gemini-2.0-flash)
• TTS: Deepgram API
• Video Processing: FFmpeg
• PDF Processing: PyMuPDF
• Frontend: HTML5, CSS3 (Neurobrutalist Design), Vanilla JS

📦 Installation

Prerequisites

• Python 3.10+
• FFmpeg installed and added to your system PATH.
• API Keys for Google Gemini and Deepgram.

Steps

1. Clone the repository:

   git clone https://github.com/yourusername/MangaNarrator.git
   cd MangaNarrator

2. Create and activate a virtual environment:

   python -m venv venv
   # Windows
   venv\Scripts\activate
   # Mac/Linux
   source venv/bin/activate

3. Install dependencies:

   pip install -r requirements.txt

4. Set up Environment Variables: Create a .env file in the root directory and add your keys:

   GOOGLE_API_KEY=your_gemini_api_key_here
   DEEPGRAM_API_KEY=your_deepgram_api_key_here
   SECRET_KEY=your_flask_secret_key

🏃‍♂️ Usage

1. Start the Backend Server:

   cd backend
   python server.py

   The server will start at http://127.0.0.1:5000.

2. Open the App: Go to http://127.0.0.1:5000 in your browser.

3. Generate a Video:

   • Upload a Manga PDF file.
   • Select a Voice Model.
   • (Optional) Tweak the "Personality System Prompt".
   • Click GENERATE VIDEO.

4. Watch & Download: Follow the real-time status logs. Once complete, the video player will appear, and a download link will be generated.

🤝 Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

📄 License

This project is open-source and available under the MIT License.