https://github.com/aneeshpatne/facerecognitionattendencesystem
Automated Face Recognition Attendance System using ML classifiers for real-time tracking through webcam with CSV logging capabilities.
https://github.com/aneeshpatne/facerecognitionattendencesystem
attendance facerecognition haar-cascade-classifier ml naive-bayes-classifier
Last synced: 2 months ago
JSON representation
Automated Face Recognition Attendance System using ML classifiers for real-time tracking through webcam with CSV logging capabilities.
- Host: GitHub
- URL: https://github.com/aneeshpatne/facerecognitionattendencesystem
- Owner: aneeshpatne
- Created: 2022-10-11T10:41:22.000Z (about 3 years ago)
- Default Branch: master
- Last Pushed: 2024-11-22T10:49:47.000Z (11 months ago)
- Last Synced: 2025-01-12T14:19:30.720Z (9 months ago)
- Topics: attendance, facerecognition, haar-cascade-classifier, ml, naive-bayes-classifier
- Language: Python
- Homepage:
- Size: 70.3 MB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# 🎯 Face Recognition Attendance System
[](https://www.python.org/)
[](https://opencv.org/)
[](https://scikit-learn.org/)
A sophisticated real-time attendance management system powered by advanced facial recognition technology. This system leverages multiple machine learning algorithms to provide accurate face detection and recognition with comprehensive attendance tracking capabilities.
## ✨ Key Features
### 🔍 Advanced Face Recognition
- **Real-time face detection** using OpenCV and webcam integration
- **High-accuracy face recognition** with face_recognition library
- **Multi-algorithm support** with performance comparison
- **Unknown face detection** with confidence thresholds
### 🤖 Machine Learning Models
- **K-Nearest Neighbors (KNN)** - Primary classifier with distance weighting
- **Support Vector Machine (SVM)** - Linear kernel with probability estimation
- **Random Forest** - Ensemble method for robust predictions
- **Naive Bayes** - Gaussian implementation for baseline comparison
### 📊 Attendance Management
- **Entry/Exit tracking** with timestamp logging
- **CSV-based records** for easy data analysis
- **Date and time stamping** with automatic formatting
- **User confirmation system** to prevent false entries
- **Duplicate prevention** and validation checks
### 🎛️ User Interface
- **Interactive webcam display** with real-time face detection boxes
- **Visual feedback** with name labels and confidence indicators
- **Console-based controls** for entry/exit selection
- **Error handling** and user-friendly prompts
## 🏗️ Project Architecture
```
FaceRecognitionAttendanceSystem/
├── 📁 classifier/ # Trained model storage
│ └── trained_knn_model.clf # Serialized KNN model
├── 📁 train_image/ # Training dataset
│ ├── person1/ # Individual person folders
│ ├── person2/ # Multiple images per person
│ └── ... # Organized by name
├── 📄 FaceRecogKnn.py # Main recognition engine
├── 📄 Train_main.py # Model training pipeline
├── 📄 attendance.csv # Attendance records
└── 📄 README.md # Documentation
```
## 🚀 Quick Start
### Prerequisites
Ensure you have Python 3.7+ installed, then install the required dependencies:
```bash
pip install opencv-python
pip install face-recognition
pip install scikit-learn
pip install pandas
pip install numpy
pip install matplotlib
```
### Installation & Setup
1. **Clone or download** this repository to your local machine
2. **Prepare training data:**
```bash
# Create individual folders for each person
mkdir train_image/john_doe
mkdir train_image/jane_smith
# Add 5-10 clear face images per person
# Images should be well-lit with clear facial features
```
3. **Train the models:**
```bash
python Train_main.py
```
This will:
- Process all training images
- Extract facial encodings
- Train multiple ML models
- Display accuracy metrics
- Save the best performing model
4. **Run the attendance system:**
```bash
python FaceRecogKnn.py
```
## 🔧 Detailed Usage
### Training New Users
1. Create a folder named with the person's full name (e.g., `john_doe`)
2. Add 5-10 high-quality face images:
- Well-lit environment
- Clear facial features
- Different angles and expressions
- Consistent lighting conditions
3. Run the training script to update models
### Recording Attendance
1. Launch the recognition system
2. Position yourself in front of the webcam
3. Wait for face detection and recognition
4. Select option:
- `1` for Entry time
- `2` for Exit time
5. Confirm the detected identity
6. Attendance is automatically logged
### Monitoring Performance
The system provides real-time feedback:
- **Green box**: Successfully recognized face
- **Red text**: Person's name with confidence
- **Console output**: Recognition status and options
## 📈 Model Performance
The system trains and evaluates four different algorithms:
| Algorithm | Typical Accuracy | Strengths |
| ----------------- | ---------------- | ------------------------------------------------- |
| **KNN** | 85-95% | Fast prediction, good with small datasets |
| **SVM** | 80-90% | Robust to outliers, works well in high dimensions |
| **Random Forest** | 75-85% | Handles overfitting, provides feature importance |
| **Naive Bayes** | 70-80% | Fast training, good baseline performance |
### Performance Optimization Tips
- **Image Quality**: Use high-resolution, well-lit images
- **Dataset Size**: Minimum 5 images per person, ideally 10+
- **Consistency**: Maintain similar lighting and camera conditions
- **Threshold Tuning**: Adjust recognition threshold based on accuracy needs
## 📊 Attendance Data Format
The system generates structured attendance records:
```csv
Name,Date,Time,Entry/Exit
john_doe,27/6/2025,09:15:32,Entry
jane_smith,27/6/2025,09:16:45,Entry
john_doe,27/6/2025,17:30:15,Exit
jane_smith,27/6/2025,17:32:08,Exit
```
### Data Fields
- **Name**: Recognized person identifier
- **Date**: DD/MM/YYYY format
- **Time**: HH:MM:SS format (24-hour)
- **Entry/Exit**: Attendance type
## 🛠️ Technical Implementation
### Core Components
#### Face Detection Pipeline
```python
# Frame processing workflow
frame → resize → color_convert → face_locations → face_encodings → prediction
```
#### Recognition Algorithm
- Uses dlib's face recognition model
- Generates 128-dimensional face encodings
- Applies distance-based matching with KNN
- Implements confidence thresholds for unknown faces
#### Training Process
- Automatically processes training images
- Extracts facial features using face_recognition
- Splits data into train/test sets (80/20)
- Trains multiple classifiers simultaneously
- Evaluates and compares model performance
## 🔧 Configuration Options
### Recognition Threshold
Adjust in `FaceRecogKnn.py`:
```python
threshold = 0.6 # Lower = more strict, Higher = more lenient
```
### Model Parameters
Customize in `Train_main.py`:
```python
n_neighbors = 2 # KNN neighbors
test_size = 0.2 # Train/test split ratio
max_depth = 2 # Random Forest depth
```
## 🐛 Troubleshooting
### Common Issues
**No face detected:**
- Ensure good lighting
- Position face clearly in camera view
- Check camera permissions
**Low accuracy:**
- Add more training images
- Improve image quality
- Retrain the model
**Recognition errors:**
- Adjust threshold values
- Add more diverse training data
- Check for similar-looking individuals
## 🤝 Contributing
1. Fork the repository
2. Create a feature branch
3. Make your improvements
4. Test thoroughly
5. Submit a pull request
## 📄 License
This project is open-source and available under the MIT License.
## 🙏 Acknowledgments
- **face_recognition** library by Adam Geitgey
- **OpenCV** for computer vision capabilities
- **scikit-learn** for machine learning algorithms
- **dlib** for facial landmark detection
---
_Built with ❤️ for efficient attendance management_