LexiAssist

A production-ready application designed to assist users with dyslexia, learning differences, visual strain, or low reading confidence. The system processes uploaded content through a pipeline of OCR, text cleaning, simplification, and voice generation.

Features

• High-Accuracy OCR: Uses Gemini 2.5 Flash to extract text from images (PNG, JPG) and PDFs.
• Text Simplification: Uses FLAN-T5 to simplify text with three severity levels (Light, Medium, Strong).
• Natural Voice Generation: Uses ElevenLabs API for high-quality, natural-sounding speech with multiple tones.
• Accessibility-First UI: High-contrast modes, dyslexia-friendly fonts, and adjustable text settings.
• History & Session Management: Save and replay previous reading sessions.

Prerequisites

• Python 3.9+
• poppler-utils (for PDF processing)
  • macOS: brew install poppler
  • Ubuntu: sudo apt-get install poppler-utils
  • Windows: Download and add to PATH

Installation

1. Clone the repository:

   git clone <repository-url>
   cd LexiAssist

2. Create and activate a virtual environment:

   python3 -m venv venv
   source venv/bin/activate

3. Install dependencies:

   pip install -r requirements.txt

4. Set up environment variables:

   • Copy .env.example to .env
   • Add your API keys:

     GOOGLE_API_KEY=your_gemini_api_key
     ELEVENLABS_API_KEY=your_elevenlabs_api_key
     ELEVENLABS_VOICE_ID=your_voice_id
     

Usage

Run the application using the virtual environment:

./venv/bin/python app.py

Open your browser to the local Gradio URL (usually http://127.0.0.1:7860).

Architecture

The application follows a linear processing pipeline:

1. Upload: User uploads an image or PDF.
2. Preprocessing: Image is cleaned (deskewed, contrast enhanced).
3. OCR: Gemini 2.5 Flash extracts text.
4. Simplification: FLAN-T5 simplifies text based on user preference.
5. TTS: ElevenLabs generates audio.
6. Storage: Session is saved to local history.

Accessibility Options

• Dyslexia-Friendly Mode: Adjusts font, spacing, and line width.
• High Contrast: Improves visibility for low-vision users.
• Voice Tones: Calm, Happy, and Neutral options for different contexts.