Tessera

App Info

Conversational voice agent for auditory rehabilitation through adaptive listening exercises. View more information here.

Features

• Real-time voice interaction with AI assistant powered by Google Gemini
• Immersive 3D audio scenarios with spatial positioning and environmental sounds
• Adaptive difficulty based on user performance and progress tracking
• Audio-based exercises for hearing rehabilitation and auditory training
• Visual feedback through animated orb interface
• Session persistence and progress monitoring

Setup

• Install Python 3.12.3 or higher
  • Confirm installation by running python --version in the command line
• Install Kivy dependencies for your platform
• Install required packages by running pip install -r requirements.txt

Build app

• Install required packages by running pip install -r requirements.txt
• From the root directory run pyinstaller Tessera.spec

API Key Setup

• Get a Google Gemini API key from Google AI Studio
• Create a .env file in the project root directory
• Add your API key to the .env file:

  GEMINI_API_KEY=your_api_key_here
  

• Privacy Note: Your API key is never saved or transmitted anywhere except directly to Google's Gemini API. It remains on your local machine only.

Development

• Run python src/main.py to start the application
• Run black src/ to format Python code

Project Structure

• src/main.py - Application entry point
• src/ui/ - Kivy-based user interface components
• src/managers/ - Core logic and state management
• src/audio_engine/ - Audio playback and mixing functionality
• src/services/ - External service integrations (Gemini, TTS, etc.)
• web/ - Static website files for GitHub Pages deployment
• audio/ - WAV audio files organized by category

License

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