This project is forked from vrnmthr/face-detecterz and aims to improve face recognition by increasing accuracy and debiasing the model. While research is still being done, changes are documented.
Run pip install -r requirements.txt
For live camera feed and recognition, run python3 repl.py and enter your name when prompted and the picture samples will be taken.
python3 repl.py --test to train and test on preset videos.
The embeddings are from the LFW dataset, which is located in embeddings/lfw