Real-Time Object Detection React App

This React application enables real-time object detection using either the device's camera or a randomly selected video from a prepared list.

Features

• Object Detection: Utilizes TensorFlow.js, a machine learning library for JavaScript, to perform object detection.
• ml5 Model: Implements ml5.js, which offers pre-trained models that can process various types of inputs (images, video, audio, text) and produce outputs such as labels and confidence scores.(based on TensorFlow.js)
• Common Objects in Context (COCO) Dataset: The model used in this app is trained on the COCO dataset, a commonly used benchmark for object detection in images.

How to Use

1. Camera Usage: Ensure your browser allows access to the camera when using the camera option.
2. Video Selection: Toggle between camera source and a random video using the provided toggle button.
3. Start/Stop Detection: Press the "Start" button to initiate object detection. Press "Stop" to halt detection.

Technologies Used

• React
• ml5.js
• COCO Dataset

Installation

To run the app locally:

1. Clone this repository.
2. Navigate to the project directory.
3. Install dependencies with npm install.
4. Start the development server with npm start.

Hosted URL

This project is hosted here