Estimate the age based on human facial images. To perform the task, we use 3 different machine learning models:

• Multinomial Logistic Regression
• Naïve Bayes classifier
• Multiclass classification with Convolutional Neural Network

1. Ensure all the required dependencies are installed.
2. Initialize the submodule for data scripts. (optional if you've already had it locally)
3. Start the client and server (order doesn't matter).

• Server

pip install --user --requirement requirements.txt

• Client

cd app

npm i

git submodule update --init

In the directory of this repository, do

python -m server.main &

cd app

npm run serve &

The datasets is composed of 187154 facial images (> 100,000 samples and > 100 features) for training the model. We combine the following 2 datasets and unify their dimensions into 250x250.

• CACD (160k)
• UTKFace (20k)

• Python 3.7+

• NodeJS 12+

• Python related: Numpy, Pandas, Tensorflow, Pytorch, CV2, PIL

• JS related: Vue, Ramda, Element-UI, Font-Awesome