This tool provides a simple method for detecting faces within an image and saving each individual face as a 512 x 512 square image
. This facilitates stable diffusion training. The tool utilizes the deepface model for face detection.
Sample jupyter notebook : extract_convert_module.ipynb
Command:
python extract_face_sq.py --input_folder='./input_img'
Extracted individual face images will be saved in the subfolder resized
of the input image folder