Exam Helper - AI-Powered Floating Assistant

A Python-based floating GUI application that works as an intelligent exam helper and personal assistant. The application captures questions from screen text (OCR) and audio input, then provides AI-powered answers using OpenAI's GPT models.

📸 Application Preview

Modern, sleek interface with dark theme and intuitive controls

✨ Key Highlights

• 🎨 Modern Dark UI: Sleek, professional interface with smooth animations
• 🤖 Multi-AI Support: OpenAI GPT, Google Gemini, and Perplexity integration
• 👻 Stealth Mode: Invisible to screen sharing (Zoom, Teams, etc.)
• 📸 Smart Capture: OCR, audio recording, and live screen monitoring
• ⚡ Real-time Processing: Instant AI responses with progress indicators

Features

🔍 Question Capture

• Screen OCR: Continuously scans the entire screen for text questions
• Audio Input: Captures questions from microphone input
• Manual Input: Text box for typing custom questions
• Smart Detection: Automatically identifies likely questions using heuristics

🤖 AI-Powered Answers

• Dual AI Integration: OpenAI for text, Perplexity for vision analysis
• Perplexity Vision: Advanced image analysis for screenshot-based questions
• Response Modes: Toggle between short/direct and detailed/explanatory answers
• Question Analysis: Automatically categorizes question types (math, definition, explanation, etc.)
• Context Awareness: Provides relevant and accurate answers

👻 Stealth Mode

• Screen Sharing Protection: Hides from screen capture applications (Zoom, Teams, etc.)
• Always Visible: Remains visible to the user while hidden from recordings
• Transparency Control: Adjustable window transparency
• Click-through Mode: Optional mouse event pass-through

⌨️ User Interface

• Modern Dark Theme: Professional appearance with smooth gradients
• Floating Window: Always-on-top, resizable interface (toggleable)
• Model Selection: Choose between different AI models for different tasks
• Quick Toggle: Ctrl+Shift+H hotkey to hide/show instantly
• Screenshot Vision: Ctrl+Shift+C hotkey for instant screenshot capture and AI analysis
• Real-time Display: Live answer updates with timestamps and status indicators
• Settings Panel: Easy configuration of API keys and preferences
• Progress Tracking: Visual feedback for all operations

🚀 Quick Start

Option 1: Download Executable (Recommended)

1. Download the latest release from the releases page
2. Extract ExamHelper.exe from the zip file
3. Run ExamHelper.exe - no installation required!
4. Configure your API keys in the settings

Option 2: Build from Source

Prerequisites

1. Python 3.8+ installed on your system
2. Tesseract OCR for text recognition
3. OpenAI API Key for AI responses

Step 1: Install Python Dependencies

python install_dependencies.py

Or manually install:

pip install -r requirements.txt

Step 2: Install Tesseract OCR

Windows:

winget install UB-Mannheim.TesseractOCR

Or download from: https://github.com/UB-Mannheim/tesseract/wiki

macOS:

brew install tesseract

Linux:

sudo apt-get install tesseract-ocr

Step 3: Configure API Keys

1. Run the application: python exam_helper.py
2. The application will create a config.json file automatically
3. Use the model selection dropdowns to choose your preferred AI models:
   • Image Recognition: OpenAI GPT-4o, Gemini Pro Vision, etc.
   • Audio Recognition: OpenAI Whisper, Gemini models
   • AI Response: OpenAI GPT models, Gemini Flash/Pro
4. Enter your API keys in the configuration file or through the interface
5. Adjust other preferences as needed

🔧 Building Executable

To create your own standalone executable:

# Method 1: Advanced build with progress tracking
python build_exe.py

# Method 2: Simple one-click build (takes long time with large application size)
build_simple.bat

# Method 3: Simple one-click build (takes less time with less application size)
build_optimize.bat

# Method 4: almost same time with same size application
build_executable.bat
build.bat

# Method 5: Manual PyInstaller
pyinstaller --onefile --windowed exam_helper.py

See BUILD_GUIDE.md for detailed build instructions.

Usage

Starting the Application

python exam_helper.py

Basic Operations

1. Automatic Scanning: The app continuously scans for questions
2. Manual Questions: Type questions in the input box
3. Screenshot Analysis: Click "📸 Capture Screen" or press Ctrl+Shift+C
4. View Answers: Responses appear in the main display area
5. Toggle Modes: Switch between short and detailed answers
6. Control Scanning: Start/Stop OCR and Audio scanning independently
7. Hide/Show: Use Ctrl+Shift+H or the Hide button
8. Always on Top: Toggle to keep window above other applications

Stealth Mode

• Automatically hides from most screen sharing applications
• Window remains visible to you but not in recordings
• Works with Zoom, Teams, Google Meet, Discord, and more

Configuration

Settings File: config.json

{
  "openai_api_key": "your-api-key-here",
  "scan_interval": 3,
  "audio_enabled": true,
  "ocr_enabled": true,
  "response_mode": "short"
}

Available Options

• scan_interval: Seconds between screen scans (default: 3)
• audio_enabled: Enable/disable microphone listening
• ocr_enabled: Enable/disable screen text capture
• response_mode: "short" for brief answers, "detailed" for explanations

📁 File Structure

exam_helper/
├── exam_helper.py          # Main application with modern GUI
├── ocr_module.py          # Screen text capture and OCR
├── audio_module.py        # Audio input and speech recognition
├── llm_module.py          # OpenAI API integration
├── gemini_module.py       # Google Gemini AI integration
├── perplexity_module.py   # Perplexity AI integration
├── stealth_module.py      # Stealth mode functionality
├── screenshot_module.py   # Screenshot capture utilities
├── config.json            # Configuration file (auto-generated)
├── requirements.txt       # Python dependencies
├── build_exe.py          # Advanced executable builder
├── build_simple.bat      # Simple build script
├── BUILD_GUIDE.md        # Detailed build instructions
├── preview.png           # Application screenshot
└── README.md             # This file

Technical Details

OCR Processing

• Uses Tesseract OCR with OpenCV preprocessing
• Applies noise reduction and image enhancement
• Filters out non-question text automatically

Audio Processing

• Real-time speech recognition using Google Speech API
• Microphone calibration for ambient noise
• Voice activity detection to reduce false positives

Stealth Implementation

• Uses Windows API calls to hide from screen capture
• SetWindowDisplayAffinity with WDA_EXCLUDEFROMCAPTURE
• Layered window attributes for transparency control

AI Integration

• Multi-Provider Support: OpenAI GPT models, Google Gemini, Perplexity
• Model Selection: Choose optimal models for different tasks
• Smart Routing: Automatic model selection based on question type
• Rate limiting: Prevents API abuse and manages costs
• Error handling: Robust network and API error recovery
• Status Indicators: Real-time model availability and health checks

Troubleshooting

Common Issues

"No OpenAI API key configured"

• Add your API key in Settings
• Ensure the key has sufficient credits

"Tesseract not found"

• Install Tesseract OCR (see installation steps)
• On Windows, may need to set path manually in ocr_module.py

"Microphone not working"

• Check microphone permissions
• Ensure microphone is not used by other applications
• Try running as administrator

"Stealth mode not working"

• Run as administrator for full functionality
• Some screen capture apps may still detect the window
• Try adjusting transparency settings

Performance Tips

• Increase scan_interval for better performance
• Disable audio or OCR if not needed
• Close other resource-intensive applications

Security & Ethics

Important Notes

• This tool is for educational assistance only
• Respect your institution's academic integrity policies
• Use responsibly and in accordance with exam rules
• The stealth feature should only be used where permitted

Privacy

• Audio is processed locally and via Google Speech API
• Questions are sent to OpenAI for processing
• No data is stored permanently by the application
• Review OpenAI's privacy policy for their data handling

Contributing

Feel free to contribute improvements:

1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Submit a pull request

License

This project is for educational purposes. Please use responsibly and in accordance with your institution's policies.

Support

For issues or questions:

1. Check the troubleshooting section
2. Review the logs in exam_helper.log
3. Ensure all dependencies are properly installed
4. Verify your OpenAI API key is valid and has credits