Vision +

Overview

The Vision + is a real-time, context-aware solution designed to assist users by integrating AI functionalities such as:

• Object detection
• Text recognition (OCR)
• Speech-to-text and text-to-speech
• Face recognition
• User feedback collection for continuous improvement

This project combines modern AI tools with a user-friendly API and PWA dashboard.

---

Features

• Real-time Object Detection: Identify and describe objects from a live feed.
• OCR: Extract and interpret text from images or camera feeds.
• Speech Interaction:
  • Convert speech to text for user queries.
  • Provide spoken responses using TTS.
• User Feedback: Collect and manage user input to improve performance.
• Dashboard: Manage saved results and session logs.

---

Tech Stack

• Backend: Node.js, Express, MongoDB
• AI Models: Python (YOLOv5, Tesseract, Whisper)
• Authentication: Google OAuth 2.0
• Deployment: Docker

---

API Endpoints

Endpoint	Method	Description
/api/detect	POST	Detect objects in an image or live feed
/api/ocr	POST	Extract text from an image
/api/ask	POST	Process a user query and provide answers
/api/speech	POST	Convert speech to text
/api/speak	POST	Convert text to speech
/api/feedback	POST	Submit user feedback

---

Getting Started

Prerequisites

• Node.js (v16+)
• MongoDB
• Docker (optional)

Installation

1. Clone the repository:

   git clone https://github.com/rabindra789/Vision_plus
   cd Vision_plus

2. Install dependencies:

   npm install

3. Set up environment variables:

   • Create a .env file with the following:

     GOOGLE_CLIENT_ID=your-client-id
     GOOGLE_CLIENT_SECRET=your-client-secret
     MONGO_URI=your-mongo-uri

4. Start the server:

   npm start

5. (Optional) Start Docker for deployment:

   docker-compose up --build

---

Usage

1. Log in using Google OAuth.
2. Access APIs via the provided endpoints.
3. Manage results on the dashboard.

---

Author

1. Rohan Bhoi
2. Swayam Kumar Sahu
3. Rabindra Kumar Meher

---

Contribution

Contributions are welcome! Please fork the repository and submit a pull request.

---

License

MIT License