Real-Time-Sign-Language-Detection-Using-Deep-Learning

Data Scientist

ML Engineer

OpenCV

Python

TensorFlow

Real-Time-Sign-Language-Detection-Using-Deep-Learning

This project demonstrates a Sign Language Detection system that utilizes Action Recognition with an LSTM (Long Short-Term Memory) Deep Learning model. The system leverages MediaPipe's Holistic Keypoints to extract skeletal data from video sequences, which is then used to predict sign language gestures.

Requirements

To run this project, you need the following dependencies:

Python (3.6 or higher)

OpenCV (cv2)

NumPy

Matplotlib

Tensorflow

Keras

Mediapipe

You can install the required libraries using pip:

pip install opencv-python numpy matplotlib tensorflow keras mediapipe

How it works

The Sign Language Detection system follows these steps:

Extract MediaPipe Holistic Keypoints: The system uses the mediapipe library to extract facial, pose, left hand, and right hand keypoints from video frames. These keypoints represent the sign language gestures.

Build a Sign Language Model: The extracted keypoints are fed into a deep neural network with LSTM layers. This model handles the sequence of keypoints, enabling effective detection of sign language actions.

Real-time Sign Language Prediction: The trained model predicts sign language gestures in real-time using video sequences. The application captures the camera feed, processes each frame, and displays the detected action label on the screen.