🩺 Face Mask Detection — Short Project Story

🌟 About the Project

A real-time AI system that detects whether a person is wearing a face mask. I built it using Python, TensorFlow/Keras, and OpenCV. The model classifies faces into with mask and without mask from webcam or image input.

🚀 Inspiration

I was inspired by the need for contactless monitoring during COVID-19 and wanted to build a practical computer vision project that helps with safety and automation.

📘 What I Learned

• How to prepare and clean image datasets
• How CNNs work for classification
• Training a model with TensorFlow/Keras
• Real-time face detection using OpenCV
• Designing a simple classification pipeline

Basic cross-entropy loss used: [ \text{Loss} = -\sum y \log(\hat{y}) ]

🛠️ How I Built It

• Organized dataset into with_mask and without_mask
• Preprocessed images (224×224, normalization)
• Trained a CNN with softmax output
• Connected model to OpenCV for live detection
• Displayed results with colored bounding boxes

⚠️ Challenges

• Cleaning low-quality dataset images
• Fixing overfitting using augmentation
• Optimizing real-time detection speed
• Installing TensorFlow + OpenCV on Windows 11

Built With

• apis
• github
• keras
• list-as-optional.)-google-colab-(for-training
• matplotlib
• numpy
• opencv
• opencv-python
• python
• sklearn
• tensorflow
• tensorflow/keras
• you-can-add-these.-for-now