search

serengil / deepface

A Lightweight Face Recognition and Facial Attribute Analysis (Age, Gender, Emotion and Race) Library for Python
https://www.youtube.com/watch?v=WnUVYQP4h44&list=PLsS_1RYmYQQFdWqxQggXHynP1rqaYXv_E&index=1
MIT License
14.29k stars 2.2k forks source link

Analyze method is legal to use for commercial purpose? #970

Closed Hassaan68 closed 9 months ago

Hassaan68 commented 9 months ago

It looks like that Age,Gender etc analysis are using VGG-Face model and VGG-Face model is not for commercial use. can we use analysis method for commercial use?

Hassaan68 commented 9 months ago

@serengil can you answer quickly please if you are available?

Hassaan68 commented 9 months ago

@sugyan @sugyan @biranchi2018 @nguyenhoan1988 @shirasael

serengil commented 9 months ago

This is already mentioned in the license section

DeepFace wraps some external face recognition models: VGG-Face, Facenet, OpenFace, DeepFace, DeepID, ArcFace, Dlib, and SFace. Besides, age, gender and race / ethnicity models were trained on the backbone of VGG-Face with transfer learning. Licence types will be inherited if you are going to use those models. Please check the license types of those models for production purposes.

Hassaan68 commented 9 months ago

@serengil . The main question is are you using the VGG-Face dataset in analyze methods or just using the architecture?

serengil commented 9 months ago

If you follow the links in readme, you will get the answer.

Detailed tutorial: https://sefiks.com/2019/02/13/apparent-age-and-gender-prediction-in-keras/

Long story short: 1- Structure of VGG-Face is being used except it final layer. 2- Weights of base VGG-Face model copied as is 3- Just the last layer's weights trained from scratch with imdb-wiki dataset

So, all layers except the last one were trained by VGG-Face dataset already.