mlkinteh2/African-Football-Player-Face-Classifier-Hakimi-Salah-Mahrez-Minteh-
Record type:
softwaremodel
Creator:
mlk
Host:
# African Football Player Face Classifier (Flask Web App)
A Flask-based web application that recognizes African football players from their facial images using **Machine Learning** and **Computer Vision**.
The app uses **OpenCV**, **Wavelet Transform**, and a **Support Vector Machine (SVM)** model to classify uploaded images of players.
---
## Overview
This project demonstrates the integration of **AI and Web Development**.
It automatically detects faces from uploaded images, extracts features using **Wavelet Transform**, and predicts the correct player using a pre-trained **SVM model**.
Users interact with a simple web interface built using **HTML, CSS, and JavaScript**, while the backend is powered by **Python Flask**.
---
## Players in the Dataset
| Player Name | Country |
|--------------|----------|
| Achraf Hakimi | π²π¦ Morocco |
| Mohamed Salah | πͺπ¬ Egypt |
| Riyad Mahrez | π©πΏ Algeria |
| Yankuba Minteh | π¬π² Gambia |
> *Note: The dataset was manually collected. Yankuba Mintehβs dataset was limited, so image augmentation techniques (flipping, rotation, brightness adjustment) were applied to increase the samples.*
---
## Machine Learning Overview
- **Face Detection:** Haar Cascade Classifier (OpenCV)
- **Feature Extraction:** Wavelet Transform (`pywt`) + raw pixel features
- **Model Used:** Support Vector Machine (SVM)
- **Framework:** scikit-learn
- **Accuracy:** ~90% (SVM performed best among tested models)
---
## π Model Performance
| Model | Accuracy | Best Parameters |
|--------|-----------|----------------|
| 𧩠SVM | **0.90** | {'C': 1, 'kernel': 'linear'} |
| π² Random Forest | 0.72 | {'n_estimators': 10} |
| π Logistic Regression | 0.84 | {'C': 1} |
**Confusion Matrix (SVM):**
---
## 𧩠Tech Stack
| Area | Technology |
|------|-------------|
| Backend | Python (Flask) |
| Frontend | HTML, CSS, JavaScript |
| Machine Learning | scikit-learn (SVM) |
| Computer Vision | OpenCV |
| Feature Extraction | PyWavelets |
| Data Handling | NumPy, β¦