Face Mask Detection using CNN & OpenCV

Overview

A deep learning project that detects whether a person is wearing a face mask or not using a CNN model built with TensorFlow/Keras and real-time detection using OpenCV.

Tech Stack

• Python
• TensorFlow / Keras
• OpenCV
• NumPy
• Matplotlib

Model Performance

• Training Accuracy: 97.7%
• Validation Accuracy: 95.2%
• Dataset: 7553 images (3725 with mask, 3828 without mask)

How to Run

Train the model:

Open and run all cells in: Face_Mask_Detection.ipynb

Real-time webcam detection:

cd Face_Mask_Detection python webcam_detection.py Press Q to quit

Results

Dataset

Face Mask Dataset - Kaggle

Notes

• Model performs best on clear frontal face images
• Performance may vary on cropped or side profile images

Author

Contributed as part of GirlScript Summer of Code 2026 (GSSoC'26)