The Sibling Detector uses a dataset (about 6 images or more per person, more is better) and uses face recognition to detect if your sibling is nearby and sounds an alert.
To make the dataset, make a folder and then making folders inside it with all the people you want (with she sibling's folder named "Sibling"). Then put 6 or more images into each folder.
Put all your pictures of your sibling in the "Sibling" folder.
This is programmed in Python 3 and has been tested in versions 3.5 and 3.6.
python pi_face_recognition.py --cascade haarcascade_frontalface_default.xml --encodings encodings.pickle
Assuming the generic human faces are in haarcascade_frontalface_default.xml and the encoded custom faces (your sibling and other people you know) are in a file called encodings.pickle.
python encode_faces.py --dataset dataset --encodings encodings.pickle --detection-method hog
Assuming the folder with the images is called dataset and the output is encodings.pickle.
Pygame (pip install pygame), imutils (pip install imutils), face_recognition (pip install face_recognition), and opencv for python (pip install opencv-python)
It could improve the face detection by letting it distinguish you and your sibling better.
It is an operating system that allows you to easily use the Sibling Detector. It runs on all Raspberry Pis supported by Raspbian Buster (it is built upon this). You can find it on the releases page on Github. It is in 2 files because Github doesn't let you have large files. A Raspberry Pi is ideal because it supports Python