Face Age Prediction

A deep learning project that predicts a person's age from facial images using a custom convolutional neural network (CNN).
This application leverages PyTorch and OpenCV for image processing and inference, trained on the UTKFace dataset.

Key Features

• Custom CNN Architecture: Implements FAPNet, a tailored CNN designed for facial age estimation.
• Human-Level Accuracy: Achieves a Mean Absolute Error (MAE) of 5.08 (train) and 5.16 (validation) — matching human-level performance.
• Trained on UTKFace: A diverse and widely used dataset for facial age prediction.
• Image Preprocessing: Uses OpenCV for robust face detection and image normalization.
• Interactive Notebooks: Includes Jupyter notebooks for exploration, training, and evaluation.
• App Interface: Easy-to-use interface for real-time prediction from uploaded images.

---

Installation

Prerequisites

• Python 3.10+
• pip

Steps

1. Clone the Repository:

git clone https://github.com/zakariaelaoufi/Face-Age-Prediction.git
cd Face-Age-Prediction

2. Create a Virtual Environment (optional):

python -m venv venv
source venv/bin/activate      # Windows: venv\Scripts\activate

3. Install Dependencies:

pip install -r requirements.txt

Usage

Run the App

python app.py

Upload a facial image and receive the predicted age in real time.

Run the Notebook

jupyter notebook

Explore the core logic in:

• Face_Age_Prediction.ipynb
• sample.ipynb

Project Structure

Face-Age-Prediction/
├── .devcontainer/            # Dev container config
├── .idea/                    # IDE configs
├── data/                     # UTKFace dataset (not included by default)
├── models/                   # Trained model checkpoints
├── Face_Age_Prediction.ipynb # Main notebook
├── sample.ipynb              # Sample testing notebook
├── app.py                    # Streamlit or Flask app (if applicable)
├── fapnet.py                 # Custom CNN model
├── image_processing.py       # Preprocessing utilities
├── requirements.txt          # Python dependencies
├── runtime.txt               # Runtime configuration
├── README.md                 # Project documentation
└── tetststts.jpg             # Sample test image

Model: FAPNet

FAPNet is a lightweight yet effective CNN designed for age regression.
It processes grayscale facial inputs and outputs a single scalar — the predicted age.

Trained using MSE loss, with a fine-tuned architecture for optimal bias-variance balance.

Performance

Metric	Value
Train MAE	5.08
Validation MAE	5.16

These results are comparable to human-level age estimation accuracy.

---

Dataset: UTKFace

The model is trained on the UTKFace dataset, a large-scale face dataset labeled by age, gender, and ethnicity. It contains over 20,000 images covering a wide age range (0–116 years), with diversity in ethnicity and gender.

Note: The dataset is not included in this repository. Please download it separately and place it in the data/ directory.

---

Contributing

Contributions are welcome!
If you find a bug, want to suggest an improvement, or add a new feature, feel free to fork this repository and submit a pull request.

1. Fork the repository
2. Create a new branch: git checkout -b feature-name
3. Commit your changes: git commit -m 'Add new feature'
4. Push to the branch: git push origin feature-name
5. Open a pull request

---

License

Feel free to use and modify it for personal or commercial use.

---

Acknowledgments

• UTKFace Dataset by Zhang et al.
• PyTorch and OpenCV
• Open-source ML research community

---

Made by Zakaria Elaoufi