About The Project

This project is a preparation stage for making a good GAN, more specifically a big dataset for face generation

you can get theoretically (Video Duration (seconds) ) * (fps) faces.

Built With

• Python 3.6
  • face_recognition 1.3.0
  • opencv-python 4.2.0.32

Getting Started

My CPU had 12 cores so I ran 10 processes simultaneously using multiprocessing however you could set the number of process by changing the value of SEGMENT

Prerequisites

install pipenv

pip3 install pipenv

Installation

1. clone this repo

   git clone https://github.com/hajindvlp/Sebom.git
   cd Sebom
   

2. install packages

   pipenv install    
   

Usage

1. Make required directories

   mkdir videos
   mkdir faces 

2. Put a video that has faces

3. Run the program

   pipenv shell
   python3 main.py

Acknowledgements

• Ageitgey's Face Recognition