This project aims to detect human faces with and without masks in real-time using a machine learning model that leverages transfer learning. The model is trained to detect face mask presence and can be used to prevent COVID-19 virus spreading.
Features Detects human faces with and without masks in real-time using a webcam feed. Uses Keras for developing the model and Resnet-50 for training. Performs data augmentation (horizontal and vertical shift, flip) as a preprocessing step. Achieves 99% accuracy while testing with hold-out images. Uses OpenCV to detect face masks in real-time. Getting Started
Clone the repository: git clone https://github.com/Aniketh999/Face-mask-detection.git
Install the required libraries: pip install -r requirements.txt
Run the maskDetection.ipynb file to train the model and test its performance
Model Training Data The training data for this model can be found in the following link: Face Mask Detection Training Data Code and Performance For the code and performance, check the maskDetection.ipynb file.