Age Estimation with Computer Vision – Good Seed

📌 Project Overview

This project helps the Good Seed supermarket chain verify customer age from photos to comply with regulations and prevent alcohol sales to minors.

A Convolutional Neural Network (CNN) was developed and trained on thousands of face images to estimate people’s ages.

🎯 Results

• The model achieved MAE ~10.13 years.
• Generalized well across different age groups and image conditions.
• Provides support for automated age verification at point of sale.

🛠️ Technologies

GlobalAveragePooling2D, Dense, Dropout, ImageDataGenerator, MobileNetV2, preprocess_input
Python, TensorFlow/Keras, CNN, ResNet50, pandas, NumPy, Matplotlib.

📊 Business Impact

The solution allows Good Seed to:

• Comply with legal regulations for alcohol sales.
• Reduce human errors in age verification.
• Improve safety and trust in the sales process.

📒 See the notebook attached