Sign Language Detection

This project demonstrates sign language detection using TensorFlow for deep learning and OpenCV for real-time image processing. The model is based on the VGG16 architecture pretrained on the ImageNet dataset.

Requirements

• Python 3.x
• Tensorflow
• opencv
• Anaconda (optional but recommended)

Installation

1. Clone the repository to your local machine: git clone https://github.com/abdu404/Sign_Language_detection.git

2. Navigate to the project directory: cd sign_language_detection

Data Preparation

1. Create a folder named SignL in the project directory.
2. Inside the SignL folder, create a folder named Data.
3. Organize your sign language data into subfolders for each class (e.g., A, B, C) within the Data folder.
4. Each subfolder should contain images of the corresponding sign.

The Dataset used

url : https://www.kaggle.com/datasets/grassknoted/asl-alphabet

Running the Code

1. Open Jupyter Notebook: jupyter notebook

2. Run the provided Jupyter Notebook cells in order:

• Cell 1: Data Preparation
• Cell 2: Model Training
• Cell 3: Model Evaluation and Saving
• Cell 4: Inference with Webcam