search

serengil / deepface

A Lightweight Face Recognition and Facial Attribute Analysis (Age, Gender, Emotion and Race) Library for Python
https://bit.ly/deepface-py
MIT License
14.68k stars 2.22k forks source link

[FEATURE]: Information about pre-trained models (e.g. FaceNet) #1252

Closed vincent-stadler closed 5 months ago

vincent-stadler commented 5 months ago

Description

Could you tell us where you acquired the models used in the deepface library? Especially the FaceNet model. Since the official model was never released by Google, there are a lot of FaceNet models in the open. It would be useful to know which one you used, i.e. which dataset it was trained on to account for possible normalisation.

Additional Info

No response

serengil commented 5 months ago

If you follow the read-me carefully, you will find the answer.

vincent-stadler commented 5 months ago

Thank you very much for the super quick response! In Sandbergs repo (https://github.com/davidsandberg/facenet/tree/master?tab=readme-ov-file#pre-trained-models) there also are multiple models, did you just take the latest one? :)

serengil commented 5 months ago

20180402-114759 -> Facenet512 20170512-110547 -> Facenet