🎧 AI AudioBook Generator

The AI AudioBook Generator is an advanced application that transforms text documents into expressive, human-like audiobooks. It extracts text from various file formats, rewrites it in a storytelling style using Gemini LLM, and converts it into natural speech using Coqui TTS or pyttsx3.

✨ Features

• 📄 Multi-Format Support: Upload PDF, DOCX, and TXT documents
• 🤖 AI-Powered Narration: Gemini LLM enhances text into audiobook-style narration
• 🎙️ Natural Speech Generation: Coqui TTS for high-quality voice synthesis
• 📴 Offline Capability: pyttsx3 fallback for offline usage
• 🔒 Secure Configuration: Environment-based API key management
• 🎨 User-Friendly UI: Clean, interactive Streamlit interface
• ⚙️ Centralized Configuration: Easy settings management via config.py

🎯 Target Audience

• Students for learning on the go
• Professionals for consuming reports and documents
• Visually impaired users for accessible content
• Content creators for repurposing written content
• Anyone who prefers listening over reading

🏗️ System Architecture

AI-AudioBook-Generator/
│
├── app.py                    # Streamlit user interface
├── config.py                 # Loads .env and manages global settings
├── llm_enrichment.py         # Gemini AI narration enhancement
├── text_extraction.py        # PDF/DOCX/TXT text extraction
├── tts_generator.py          # Coqui TTS + pyttsx3 audio generation
├── requirements.txt          # Python dependencies
├── .env.example              # Environment variables template
└── README.md                 # Project documentation

📋 Prerequisites

• Python 3.11 or higher
• Git
• Internet connection (for Gemini API and Coqui TTS)
• API keys for Gemini AI

🚀 Quick Installation

1️⃣ Clone the Repository

git clone https://github.com/Harsha-2005/AI-AudioBook-Generator.git
cd AI-AudioBook-Generator

2️⃣ Create Virtual Environment

# Windows
python -m venv venv
venv\Scripts\activate

# macOS/Linux
python -m venv venv
source venv/bin/activate

3️⃣ Install Dependencies

pip install -r requirements.txt

4️⃣ System Dependencies

Windows:

• Install eSpeak NG
• Add it to PATH
• Restart terminal

Ubuntu/Linux:

sudo apt update
sudo apt install espeak-ng

macOS:

brew install espeak

🔐 Configuration

1. Get API Keys

1. Visit Google AI Studio
2. Generate a Gemini API key
3. (Optional) Get OpenAI API key if needed

2. Configure Environment

Create a .env file in the project root:

# API Keys
GEMINI_API_KEY=your_gemini_api_key_here
OPENAI_API_KEY=your_openai_api_key_here

# TTS Configuration
TTS_ENGINE=coqui
TTS_OUTPUT_FORMAT=wav

# Application Settings
DEBUG_MODE=False

⚠️ Important: Add .env to .gitignore to keep your keys secure!

🖥️ Usage

Start the Application

streamlit run app.py

Using the Web Interface

1. Upload Document: Drag and drop or select a PDF, DOCX, or TXT file
2. Preview Text: Review the extracted text before processing
3. Generate Audiobook: Click the "Generate Audiobook" button
4. Listen/Download: Play the audio directly or download the generated file

Workflow

Upload Document → Extract Text → AI Narration Enhancement → Convert to Speech → Download Audiobook

🧠 How It Works

1. Text Extraction

• PDF: Uses PyPDF2 & pdfplumber for accurate text extraction
• DOCX: Leverages python-docx for Word document parsing
• TXT: Direct file reading with encoding detection

2. AI Narration Enhancement

• Gemini LLM rewrites text into audiobook-style narration
• Intelligent chunking prevents token overflow
• Adds expressive elements for better listening experience

3. Text-to-Speech Generation

• Primary: Coqui TTS for natural, human-like speech
• Fallback: pyttsx3 for offline functionality
• Configurable output formats (WAV, MP3)

4. User Interface

• Streamlit-based interactive UI
• Real-time progress tracking
• Audio playback and download options

🧪 Testing

Unit Testing

python -m pytest tests/

Test Coverage

• ✅ Text extraction from all supported formats
• ✅ LLM narration enhancement
• ✅ Audio synthesis with both TTS engines
• ✅ Error handling and fallback mechanisms

🚀 Deployment Options

Option 1: Streamlit Cloud

1. Push code to GitHub
2. Connect to Streamlit Cloud
3. Add environment variables in settings

Option 2: Hugging Face Spaces

1. Create new Space
2. Select Streamlit SDK
3. Upload code and configure secrets

Option 3: Docker

# Build Docker image
docker build -t audiobook-generator .

# Run container
docker run -p 8501:8501 audiobook-generator

Option 4: Local Deployment

# Run as background service
nohup streamlit run app.py --server.port 8501 &

🧰 Tech Stack

Category	Technology
Programming Language	Python 3.11
UI Framework	Streamlit
AI/ML	Google Gemini LLM
Speech Synthesis	Coqui TTS, pyttsx3
Text Extraction	PyPDF2, pdfplumber, python-docx
Configuration	python-dotenv
Document Parsing	PyMuPDF, docx2txt

🔧 Troubleshooting

Common Issues

1. API Key Errors

   • Verify .env file exists and contains correct keys
   • Check API key validity at Google AI Studio

2. TTS Engine Issues

   • Ensure eSpeak NG is properly installed
   • Check internet connection for Coqui TTS

3. Memory Issues

   • Reduce chunk size in config.py for large documents
   • Close other applications to free up memory

Logs

Enable debug mode in .env for detailed logging:

DEBUG_MODE=True

📈 Performance Metrics

Document Size	Processing Time	Audio Duration
1-10 pages	1-2 minutes	5-15 minutes
10-50 pages	3-5 minutes	15-60 minutes
50+ pages	5-10+ minutes	60+ minutes

🔮 Future Enhancements

• 🎭 Multi-voice selection for different characters
• 🌍 Multi-language support for global accessibility
• 🎵 Background music mixing options
• 📖 Chapter-wise audio segmentation
• ☁️ Cloud storage integration for saving audiobooks
• 👥 User authentication and library management
• 📱 Mobile app development
• 🔍 OCR support for scanned documents

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

🙏 Acknowledgments

• Google Gemini AI
• Coqui TTS
• Streamlit
• All open-source libraries used in this project

👨‍💻 Author

Harsha Pavan Maddala

• GitHub: @Harsha-2005
• LinkedIn: Harsha Pavan Maddala