/eGreenBinAI_FaceRecognition

Primary LanguagePythonMIT LicenseMIT

Real time face recognition

Detect and recognize face in realtime.

Demo

Run local

Expect Python >= 3.10

  1. Install dependencies
pip install -r requirements.txt
  1. Prepare data

There are two options

  • Manually creating facebank data as:
/facebank
├── /name1
│   ├── *.jpg
└── /name2
    ├── *.jpg
  • Get faces from webcam by running:
python make_data.py
  1. Create face embedding & train a classifier

Create face embedding for each image in the dataset. Expect 1 face each images. Embedding from ArcFace is a vector (1,512).

python make_identifier.py
  1. Inference

Run the following command for details information:

python main.py --help

  • -t tag for task, img for inference on image, vid for inference on video.

  • -p path to the image file.

  • -s save the image/video after processing.

Run API

Run on local

bash run_api.sh

Docker

  1. Build the container
bash build.sh
  1. Run the container
bash run.sh
  1. Run API
bash run_api.sh