Skip to content

Latest commit

 

History

History
135 lines (87 loc) · 5.6 KB

README.md

File metadata and controls

135 lines (87 loc) · 5.6 KB
Raw

Realtime Sign Language Translation

A real-time application for detecting and recognizing sign language gestures using a webcam feed.

Table of Contents

Aim

This project aims to create a sign language translator using machine learning techniques and Python programming. The application utilizes various modules, primarily Mediapipe, Landmark, and Random Forest algorithms to interpret and translate sign language gestures into text or spoken language.

Project Overview

Sign language is a crucial form of communication for individuals with hearing impairments. This project focuses on bridging the communication gap by creating a tool that can interpret sign language gestures in real-time and convert them into understandable text or speech.

This project leverages Flask for the web interface and TensorFlow/Keras for the machine learning model to recognize sign language gestures in real-time from a webcam feed.

American Sign Language Convention for Alphabets.

Custom Sign Language for Words / Sentences.

Demo

Showcasing a demonstration of the Realtime Sign Language Detection 

sign2text_demo.mp4

Features

  • Real-time sign language recognition: Captures hand gestures using the Mediapipe library to track landmarks and movements.
  • Landmark analysis: Utilizes Landmark module to extract key points and gestures from hand movements.
  • Machine learning translation: Employs Random Forest algorithm to classify and interpret gestures into corresponding text.
  • Text-to-speech: For better communication the text can be converted to spoken language using the speech synthesis.

Getting Started

To get started with the Sign Language Translator, follow these steps:

Prerequisites

  1. Python: Provides a vast array of libraries and frameworks for machine learning, computer vision, and data processing.
  2. TensorFlow: For building and training machine learning models.
  3. Scikit-learn: For implementing the Random Forest algorithm for sign language recognition.
  4. Numpy: For numerical computations and data manipulation.
  5. Mediapipe: For real-time hand tracking and landmark detection.
  6. OpenCV: For video processing and computer vision tasks.
  7. Flask: Web framework to develop the application.
  8. Flask-SocketIO: Adds low-latency bi-directional communication between clients and the server to Flask applications.

Installation

  1. Clone the repository:
git clone https://github.com/uzibytes/sign2text.git
cd sign2text
  1. Create and activate a virtual environment:
python -m venv venv
source venv/bin/activate  # On Windows use `venv\Scripts\activate`
  1. Install required libraries:
pip install -r requirements.txt
  1. Ensure a webcam is connected to your system.

Usage

  1. Start the Flask application:

    python app.py

  2. Open your web browser and navigate to :

     http://127.0.0.1:5000/

  3. The web interface will display the webcam feed and detected sign language gestures.

Project Report

For detailed insights, analysis, and findings, refer to the Project Report provided in the repository or click here.

Contributing

Contributions are welcome! If you'd like to contribute to this project, feel free to open issues, create pull requests, or reach out to discuss potential improvements.

License

This section states that the Realtime Sign Language Detection Using LSTM Model project is released under the MIT License. It briefly describes the terms and conditions of the license, such as the permission to use, modify, and distribute the project, with appropriate attribution. It provides a link to the full text of the MIT License for further reference.

Contact

This is a Final Year B.Tech Project for the session 2020-24. This project is completed under the Guidance of Dr. Shashi Raj (Assistant Professor, Dept. of CSE, Bakhtiyarpur College of Engineering, Patna). This is a group project and the members are :

  1. Ujjwal Raj - 20105126034 Connect on LinkedIn
  2. Krishna Raj - 20105126040 Connect on LinkedIn
  3. Prashant Kumar - 20105126043 Connect on LinkedIn
  4. Rajnish Puri - 20105126031 Connect on LinkedIn

For any questions or inquiries, feel free to contact.