Face-Recognizer

Detect specific faces from a videostream and save unknown faces using facial recognition.

To use:

1. git clone https://github.com/darshg321/Face-Recognizer.git
2. cd face-recognizer
3. pip install -r requirements.txt
4. python main.py

Green box around a face means it's in sample_images, red means it's new Known faces will be saved with name of the original face in sample_images + number, unknown faces will be saved with unknown + number

Args:

--images_path - Path to the sample images folder

--video_path - Video file. Supported formats: .mkv, .mp4, .avi, .mov, .wmv

--min_probability - Minimum probability for a face to be detected. Default is 0.95, max is 1. Higher values mean more strict detection.

--max_distance - Threshold for the distance of faces during comparison. Lower values are more strict comparing, with a max of 1. Default is 0.4