Face Mask Detection Project 😷

📌 Overview

This project is a Computer Vision pipeline for:

1. Detection: Using YOLO to detect people in a video and crop their faces.
2. Classification: A PyTorch model trained on a Kaggle dataset to classify if a person is:
   • ✅ With Mask (label = 1)
   • ❌ Without Mask (label = 0)

The project saves cropped faces into tmp/files/, then predicts mask usage and saves results into a .csv file.

---

📂 Dataset Structure

The dataset is from Kaggle: Face Mask Dataset

Organized as:

dataset/ ├── train/ │ ├── with_mask/ │ └── without_mask/ └── test/ ├── with_mask/ └── without_mask/

yaml Copy code

---

⚙️ Requirements

Install these libraries before running:

pip install torch torchvision
pip install opencv-python
pip install pandas
pip install matplotlib
pip install ultralytics
pip install pillow
🚀 How to Run
Download the dataset from Kaggle and place it in the dataset/ folder.

Run the Jupyter Notebook step by step:

Detection (YOLO): Detect people in a video → crop → save in tmp/files/.

Classification (PyTorch): Train classifier on dataset → predict on tmp/files/.

Save predictions into predictions.csv.

Check your results in the outputs/ folder.

📊 Results
Accuracy achieved: XX% (replace with your result after training).

Example detections:

With Mask ✅	Without Mask ❌

📦 Project Structure
bash
Copy code
Computer Vision Project/
│
├── dataset/              # Kaggle dataset
├── tmp/files/            # Cropped faces from YOLO
├── outputs/              # Results (plots, sample predictions, csv)
├── detection.ipynb       # YOLO detection notebook
├── classification.ipynb  # Mask classification notebook
├── requirements.txt      # Python dependencies
└── README.md             # Project documentation