Real-Time Hand Sign Detection

Welcome to the Real-Time Hand Sign Detection repository! This project leverages deep learning to detect and recognize various hand signs in real-time, using a custom-trained YOLOX model. The system is designed for applications such as gesture-based interaction, sign language interpretation, and more.

Table of Contents

• Features
• Demo
• Installation
• Usage
• Project Structure
• Model Training
• Auto Labeling Tool
• Future Work
• Contributing
• License

---

Features

• Real-Time Detection: Uses a YOLOX model for accurate, real-time hand sign detection.
• Customizable and Scalable: Includes utilities for custom data labeling, model training, and evaluation.
• Auto Labeling Tool: A built-in tool for automatic data labeling to streamline the preparation of training datasets.
• Python-based Implementation: Easily customizable and extendable Python code.

---

Demo

A demo of the hand sign detection in action can be run by following the instructions below. The demo uses the Ninjutsu_demo.py script for detecting hand signs in a live video feed.

---

Installation

To set up this project locally, follow these steps:

1. Clone the Repository

   git clone https://github.com/Chowdhurynaseeh/Realtime-handsign-detection.git
   cd Realtime-handsign-detection

2. Install Dependencies Make sure to have Python 3.7+ installed. Then, install the required dependencies:

   pip install -r requirements.txt

3. Download Model Weights Download the pre-trained YOLOX model weights from YOLOX GitHub and place them in the model/yolox directory.

---

Usage

To run the real-time hand sign detection demo:

python Ninjutsu_demo.py