jpwhite3/ai-face-extractor

A command line program that detects faces in photos and makes thumbnail groups of each unique face

ai-face-extractor

A command line program that detects faces in photos and makes thumbnail groups of each unique face. This can be helpful when gathering photos for Stable Diffusion model training.

Setup

Install dependencies

pipenv install

Run

pipenv run python -m extract_faces /path/to/photos

This will create a faces sub-directory in the provided photos directory, which will have group# sub-directories within it. Each group# directory will contain thumbnails of the extracted faces, grouped by similarity.

Known Issue: Grouped photos are a little hit-and-miss at the moment. In a future version, furture similarify measurements will be taken into account to group photos more accuratly.

Help

usage: FaceExtractor [-h] [--model {VGG-Face,Facenet,Facenet512,OpenFace,DeepFace,DeepID,ArcFace,Dlib,SFace}] [--size SIZE] source_dir

Extracts faces from images in provided source directory

positional arguments:
  source_dir            Directory containing source images

options:
  -h, --help            show this help message and exit
  --model {VGG-Face,Facenet,Facenet512,OpenFace,DeepFace,DeepID,ArcFace,Dlib,SFace}
                        Model to use (default: Facenet512)
  --size SIZE           Resolution of saved faces (default: 224)