Adding new people in a classifier

ageitgey / face_recognition

The world's simplest facial recognition api for Python and the command line

MIT License

52.94k stars 13.44k forks source link

Adding new people in a classifier #493

Open Liuxi-G opened 6 years ago

Liuxi-G commented 6 years ago

face_recognition version:
Python version:3.5
Operating System:windows

Description

Hi! I'm right now using your method (KNN) to recognise people in a large dataset, my problem is if I have already trained a classieier (for example Facescrub), how can I add new people to recognise without training the part that has been trained already? Thank you !

What I Did

Paste the command(s) you ran and the output.
If there was a crash, please include the traceback here.

ageitgey commented 6 years ago

What are you trying to do exactly? Why do you need all the Facescrub faces in your model?

Liuxi-G commented 6 years ago

I'm trying to recognize as much face as I could in a large database

cftang0827 commented 6 years ago

I don't think you need a knn classifier, you can just use L-2 norm distance to justify whether two faces are the same.

Besides, if you use KNN method, you need to retrain the classifier model as long as you add a new person.

Thanks

xSNYPSx commented 5 years ago

https://github.com/ageitgey/face_recognition/issues/787