Goal

Training a model that will differentiate between people wearing mask or not.

Dataset description

The dataset is available on Kaggle. The dataset consists of 600 training images, 306 validation images, and 100 for testing.

Training Methodology

In this project, I used only transfer learning.

• Load Data
• Data augmentation
• Model Construction : EfficientNetB1
• Model Training

Results

At the end of the training phase, I got 0.9869 for the validation accuracy and 0.0486 for the validation loss.

Testing

How to use it ?

• Install tensorflow
• install open cv
• clone repository git clone git@github.com:Ahmed-Camara/Face-Mask-Detection.git
• execute python3 prediction.py in the terminal.