Skip to content

Latest commit

 

History

History
89 lines (64 loc) · 3.51 KB

README.md

File metadata and controls

89 lines (64 loc) · 3.51 KB
Raw

MultiModelFaceAntiSpoof

🚀 MultiModelFaceAntispoof utilizes Multiple Open Source Spoofing models for Inference via Flask API. It integrates multiple face anti-spoofing models from GitHub. Contributions, corrections, and requests are welcomed on GitHub.

Build status PyPI version Docker Passing GitHub Actions Docker Compose Python TensorFlow PyTorch MTCNN Flask Gunicorn Nginx Git LFS

MultiModelFaceAntiSpoof is available on Docker Hub.

Architecture

Architecture

Installation

Setup

sudo apt install  git-lfs -y
git  clone https://github.com/sanjay7178/MultiModelFaceAntiSpoof
cd MultiModelFaceAntiSpoof
git-lfs pull

Inference

docker-compose  up  --build -d

Try Out on Google Colab


🔬 Spoof Detection Models:

  1. M1: HyperFAS
  2. M2: Face Anti-Spoofing
  3. M3: CelebA-Spoof
  4. M4: Awesome Face Antispoofing
  5. M5: Simple CNN model (trained from scratch)
  6. M6: Silent-Face-Anti-Spoofing
  7. M7: Face-Anti-Spoofing-using-DeePixBiS
  8. M8: FaceAntiSpoofing

Requirements:

  • Python 3.6 or higher
  • Upgrade pip: pip3 install pip --upgrade
  • Install required packages: pip3 install -r requirements.txt

Usage:

  • Run the test script: python3 test.py

TODO:

  1. Add more spoof detection models

Credits and Source

Original repository: FaceAntiSpoofing