/Masked-Face-Recognition

Identify subjects from CCTV footages even when there are only limited number of photographs available for the suspected individuals. The system should work in near real time and even when there are partial occlusion, different pose, and physical changes like beard , eye-glasses and face masks .

Primary LanguageJupyter Notebook

Masked Face Recognition

Alt Text

We should be able to identify subjects from CCTV footages even when there are only limited number of photographs available for the suspected individuals. The system should work in near real time and even when there are partial occlusion, different pose, and physical changes like beard , eye-glasses and face masks .

Deployment

To deploy this project there are 2 options :

  • Run on COLAB : Open Colab folder in the repo
  • Run Locally : Coming Soon

Prerequisites

Install these libraries

Features

  • Recognition at varying angles varying angle
  • Recognition of people wearing surgical masks due to covid-19 mask face
  • Identifying using limited suspect images
  • No training requied to be done on dataset (dynamic recognition)

Contributing

Contributions are always welcome!

Contributors

portfolio portfolio portfolio portfolio