AI Sign Language Project

DEMO

Below is a basic demo of the project using ASL! Enjoy <3

Overview

This project aims to recognize and translate sign languages using a CNN model (LSTM). It leverages TensorFlow, Mediapipe, and NLP for feature extraction, model training, and text correction.

Setup Instructions

1. Clone the Repository

Clone the repository to your local machine using Git:

git clone https://github.com/godinezsteven1/AI-SignLanguage.git
cd AI-SignLanguage

2. Create a Virtual Environment

Create a virtual environment to manage project dependencies:

python3 -m venv fai_project_env

Activate the virtual environment:

• For macOS/Linux:

  source fai_project_env/bin/activate

• For Windows:

  fai_project_env\Scripts\activate

3. Install Dependencies

Install the required Python packages using the requirements.txt file:

pip install -r requirements.txt

4. Verify Setup

To make sure everything is working, try running the hand_tracking.py script to ensure TensorFlow and Mediapipe are installed correctly.

cd scripts
python hand_tracking.py

If everything is set up correctly, you should see a window showing hand tracking in real time.

Running the Project

In the root directory:

python sign_recognition_gui.py

---

Environment Variables Setup -- Runing the Reddit Scrapper on your own!

This project uses environment variables for sensitive configuration. A template file .env.example is provided in the repository. Follow these steps to create your own .env file: Where CLIENT_ID, CLIENT_SECRET, and USER_AGENT is found at 'https://www.reddit.com/prefs/apps' by creating your own API app key under YOUR account.

1. Copy the template file:

   cp .env.example .env

2. Fill in information:

CLIENT_ID=your_client_id_here
CLIENT_SECRET=your_client_secret_here
USER_AGENT=your_user_agent_here
POST_LIMIT=NUMBER_LIMIT_HERE

In order to get this information please go on Reddit, click on your avatar → User Settings → scroll down to the "Apps" section. You can manage and create apps from there.

Results

The model successfully recognizes and classifies signs in real-time across multiple sign languages.

Sign Language Recognition Output

American Sign Language (ASL)

German Sign Language (DGS)

Spanish Sign Language (LSE)