This is a 4th year group project by Group F submitted to Mr Mokodir, Decision Support Systems Lecturer.
Objective: To build a face recognition program that does not gender assasinate but detects the gender of a person just by scanning their face
This are the female and male subfolders in the datasets directory containing more than 100 pictures of female and male face pictures in each respective subdirectory.
This project classifies faces according to their gender; either Female or Male. Convolutional Neural Networks are used to build the model. It uses opencv-python to video capture and the gender is predicted by the model. Opencv allows users to perform image processing and vision tasks. A squarebox is displayed predicting if either the person is female or male. By showing the camera a picture, the program will also predict if either the picture displayed is male or female.
Clone the repository and change directory into the folder:
git clone https://github.com/wagura-droid/Facial-Recognition-by-Infomatrix.git
cd Facial-Recognition-by-Infomatrix
Create an empty directory called training. This is where the models generated and saved will be stored.
mkdir training
Install virtual environment: Linux users:
sudo apt install python3.10-venv
Windows users:
pip3 install virtualenv
Create virtual environment: Linux users:
python3 -m venv <virtualenvname>
Windows users:
virtualenv <virtualenvname>
Activate virtual environment: Linux users:
source <virtualenvname>/bin/activate
Windows users:
cd <virtualenvname>/Scripts
activate
cd ../..
Install Dependencies:
pip3 install -r requirements.txt
Run the program
python3 GenderClassification.py
Inside the screenshots folder.