AI Camera with Object Detection and Audio Feedback

This project leverages YOLOv8 for object detection, EasyOCR for text recognition, and the Gemini API for scene description. It provides real-time audio feedback on the position of a person within a video frame to help them adjust their position while recording.

Features:

• Person Detection: Uses YOLOv8 to detect a person in the frame.
• Position Feedback: Provides audio feedback about the person's position (e.g., "left", "center", "top") to help them adjust.
• Text Recognition: Recognizes and reads out any text detected in the frame using EasyOCR.
• Real-time Scene Description: Uses the Gemini API to generate a description of the scene including the person’s position and any detected text.

Requirements:

Before running the app, make sure you have the following dependencies installed:

• Python 3.7 or higher
• YOLOv8 (ultralytics)
• EasyOCR
• pyttsx3
• Gemini API (Google)
• OpenCV
• numpy

Install dependencies:

To set up the environment, open the terminal and run the following commands:

# Install required libraries
pip install opencv-python pyttsx3 easyocr google-generativeai ultralytics numpy

Setup:

1. Download YOLOv8 Weights:

   • Download the YOLOv8 weights (yolov8n.pt) from the official YOLO website.
   • Place the weights in the same directory as the Python script or specify the path to it in the code.

2. Gemini API Key:

   • Replace the GEMINI_API_KEY in the code with your actual Gemini API key.

   GEMINI_API_KEY = "your_api_key_here"

Running the Application:

Once the dependencies are installed and the necessary files are set up, run the following command in your terminal to start the AI camera app:

python ai_camera.py

• The application will open a webcam feed and start detecting objects and text in the frame.
• If a person is detected, it will give audio feedback about their position (e.g., "The person is in the center of the frame").
• If text is detected, it will also be read aloud.
• Press q to exit the application.

Usage:

• Position Feedback: When the app detects a person in the frame, it will tell you where they are, like "top left", "center", or "bottom right".
• Text Recognition: If any text appears in the frame, the app will read it out loud.
• Scene Description: The app uses the Gemini API to generate a scene description, including the position of the person and any detected text.

Troubleshooting:

• Webcam not working: Ensure your webcam is properly connected and accessible. If you're using a virtual environment, check if the webcam is recognized.
• Text not detected: Make sure the text in the frame is clear and legible. The OCR might struggle with poorly lit or blurry text.
• API errors: Double-check your Gemini API key. If you encounter issues with the Gemini API, ensure your key is valid and not expired.

License:

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

Acknowledgments:

• YOLOv8: Used for real-time object detection.
• EasyOCR: Used for text recognition in images.
• Gemini API: Used to generate scene descriptions.
• pyttsx3: Used for text-to-speech conversion.

---

Feel free to fork or clone this repository and contribute to its improvement!

### Instructions:
1. **Install Dependencies**: Run the `pip install` command to install all the required Python libraries.
2. **Set Up YOLO Weights**: Make sure you have the `yolov8n.pt` weights in your project directory or update the path in the code.
3. **Replace Gemini API Key**: Replace the placeholder `your_api_key_here` with your actual Gemini API key.
4. **Run the App**: Use the terminal to run the Python script and start the real-time AI camera.