Real-Time-Sign-Language-Detection

Welcome to the GitHub repository for our ASL Detection System, developed as part of the "Introduction to Machine Learning (IML)" course. This project utilizes a convolutional neural network (CNN) to recognize American Sign Language (ASL) gestures in real-time, enhancing communication for the deaf and hard-of-hearing community.

Project Team

• Adoksh M Bhardwaj
• Ajay S Prakash
• K Jaidev Shankar Reddy
• Manvitha R Kabbathi

Project Structure

This repository is organized into several sections:

1. Data Collection
2. Data Preparation
3. Model Training
4. Real-Time Detection
5. Dataset

Step-by-Step Guide

1. Data Collection

To collect the ASL gesture images, use the collectdata.py script. This script utilizes a webcam to capture images of ASL gestures when a specific key corresponding to each letter is pressed. Ensure you have a webcam connected and operational before starting this step.

Usage:

python collectdata.py

2. Data Preparation

Once data collection is complete, organize the data into training and validation sets using the split.py script. The data is split in the ratio of 0.8 for training and 0.2 for validation. This will help in training the model effectively by providing a set of unseen images for validating the model's performance.

Usage:

python split.py

Make sure to paste the training and validation paths in the trainmodel-2.ipynb for proper model training.

3. Model Training

Train the CNN using the trainmodel-2.ipynb Jupyter notebook. This notebook outlines the structure of the CNN and contains code for training the model on the collected data. After training, the model architecture is saved as a JSON file and the model weights as an H5 file. Ensure you have Jupyter Notebook installed to open and run this file.

Usage: Open and run the notebook in Jupyter:

jupyter notebook trainmodel-2.ipynb

4. Real-Time Detection

After training the model, use the realtimedetection.py script to start the real-time detection of ASL gestures. This script loads the trained model using the JSON file and H5 file paths provided. It uses the webcam to detect and classify ASL gestures live.

Usage:

python realtimedetection.py

5. Dataset

The dataset used for training and validation, consisting of 4,466 ASL gesture images, is included in this repository. It is divided into training and validation sets, with detailed organization in the dataset directory.

Requirements

• Python 3.8+
• Libraries: OpenCV, TensorFlow, Keras, NumPy
• Jupyter Notebook or Google Colab for running .ipynb files

Installation

Clone this repository to your local machine:

git clone Repository Link 

Install the required libraries:

pip install -r requirements.txt

Thank You