/teknofest23-sign-app

Sign Language Application

Primary LanguagePythonMIT LicenseMIT

Sign-Language-Application

Description

  • This is a sign language application with features that helps disabled people to comminucate with others.

Our Motivations

  • Application motivations ;
    • Sign Language: To be able to make it widespread
    • Sign Language: Creating Awareness
    • Sign Language: Making it easy to use
    • Sign Language: Teaching efectively

Features

  • Application constructed %100 python structure with Python Language and pythonist modules.
  • Application also able to run predictions though the KeyMaps shown in ../Assets/Readme/Key Map.png file. Saved for future use.
  • Application provides features such as ;
    • Speech to text
    • Text to speech
    • Text Translation
    • Speech Recognition
    • Speech Synthesis
    • Image Recognition
    • Sign Language Recognition
    • Sign Language Synthesis
    • Graphical User Interface

Structure

  • Application is divided into two parts ;
    • Detecting Sign Language
    • Detecting Speech
  • Application is seperated into utilitiy methodsi that runs down the program much more effectively with threads and processes. Some of the utility method used in the application are ;
    • @translate_text
    • @vocalize_text
    • @listen_text
    • @detect
  • Application is constructed with a dynamical structure that can be easily modified and updated with only one line change for all other thanks to ../Constants/path_track.py file.
  • Application is served with tkinter library for graphical user interface.

Installation

Application DOES requires CUDA and CUDNN to be installed on your computer ! From scratch you can follow installation steps, that also includes CUDA installation.
You can also see the specific quedlines via @"Wired Installation.txt" from ../Assets/Readme/ folder.
Application is reserved on a machine with RTX 3060 Laptop GPU and python 3.10.0 version.\

  • Required Pre-Installation Steps
    • install visual studio 2022 via website
    • install the cuda12 from nvidia installer
  • Environment Setup
    • conda create -n sign_app python==3.10.0
    • conda activate sign_app
  • PIP and Conda installation Steps
    • conda install -c conda-forge cudatoolkit=11.2 cudnn=8.1.0
    • python -m pip install "tensorflow<2.11"
    • pip install mediapipe==0.9.0.1
    • pip install opencv_python-4.7.0.68
    • pip install playsound==1.2.2
    • pip install SpeechRecognition==3.9.0
    • pip install googletrans==3.1.0a0
    • pip install gTTS==2.3.1
    • pip install pyaudio==0.2.13
    • pip install pygrabber==0.1

Usage

  • Application is ready to use after installation steps.

  • After the "Main Menu" window the two Applications will be available to use in order.

  • To Run :

    • conda activate sign_app
    • python main.py

Example Of Usage

Example Of Usage

TO:DO

  • Enable World-Wide language support (Structure is ready, just implement it)
  • Enable for different types of sign languages (Structure is ready, just implement it)
  • Collect More data for model.
  • Collect More animation gifs.