Gender Classification

CNN classifier to predict age and gender of a face.

Dependencies:

• Numpy
• PIL
• Matplotlib
• Random
• Tensorflow
• Keras
• CV2

Installation:

1. Clone the genderClassification repository to your local machine using git clone https://github.com/benekeneb/genderClassification.git
2. In order to train the CNN you need to download the dataset UTKFace (Source: https://susanqq.github.io/UTKFace/). The easiest way to do that is to clone the following repository into your local genderClassification folder. git clone https://github.com/ongteckwu/utkface.git
3. Create virtual environment with the modules I listed above.

Project description

• webcam.py: Script which uses your webcam to classify visible faces. Start the webcam.py in the main directory to predict both age and gender. There are also webcam.py scripts in the age_classification and the gender_classification directories to just predict one thing.
• evaluate_train.py: Script to plot some example images from the dataset with the associated predicted labels by my models. Can also be found in the age_classification and the gender_classification directories. Example is evaluate_train_plot.png
• plot_histograms.py: Script to plot histograms of the used dataset.
• To train the two models there are train.py scripts in both the age_classification and the gender_classification directories.