Real-time sign language detection using CNN and MediaPipe for hand landmark recognition. Implements deep learning models to classify sign language gestures from live video input.
Real-Time Sign Language Detection
This project implements a real-time sign language detection system using a Convolutional Neural Network (CNN) and MediaPipe for hand landmark detection. The system captures live video input, processes hand gestures, and classifies them into corresponding sign language alphabets.
Project Structure:
CNNModel.py
: Defines the Convolutional Neural Network (CNN) architecture used for classifying hand gestures.handLandMarks.py
: Handles the detection of hand landmarks using MediaPipe and processes them for use by the CNN model.mediapipeHandDetection.py
: Integrates MediaPipe to perform real-time hand detection through the webcam.realTime.py
: The main script that ties everything together, using the CNN model and MediaPipe for real-time sign language detection.training.py
: Script used for training the CNN model on a dataset of hand gestures.testCNN.py
: Script for testing the performance of the trained CNN model on a test dataset.CNN_model_alphabet_SIBI.pth
: Pre-trained CNN model weights used for classification.How to Run the Project:
Make sure you have Python installed on your system. You can install the required Python packages using pip:
pip install -r requirements.txt
If you don't have a requirements.txt
file, you can manually install the necessary packages:
pip install opencv-python mediapipe torch numpy pandas
To start the real-time sign language detection, run the following command:
python realTime.py
This will activate your webcam and start detecting and classifying hand gestures in real-time.
If you want to train the CNN model from scratch, you can run:
python training.py
This script will use a dataset of hand gestures to train the model.
To test the performance of the trained CNN model on a test dataset, you can run:
python testCNN.py
How It Works:
Hand Landmark Detection:
Feature Extraction:
Gesture Classification:
Real-Time Feedback:
Requirements:
Contributing:
Contributions are welcome! If you have any ideas, suggestions, or improvements, feel free to open an issue or submit a pull request.
Contact:
For any questions or suggestions, please feel free to contact me at [[email protected]].