search

seasonSH / DocFace

Face recognition system for ID photos
MIT License
371 stars 124 forks source link

the weight imprinting implement question #8

Closed ronghuaiyang closed 5 years ago

ronghuaiyang commented 5 years ago

thanks for the excellent work ! I try to implement your work in pytorch. When I implement the weight imprinting, i set the learning rate of the classifier layer to 0. then i use the follow steps: 1、calculate the features and normalize them as the weight vecotors. 2、replace the old weight vector 3、calculate the loss use am softmax 4、backward gradients and update weights except the classifier layer.

is it correct ? i am not familiar with the tensorflow , so i don't understand your code very well. please give me some advices, thanks !

seasonSH commented 5 years ago

Yes, they are correct. Make sure you normalize twice: (1) normalize the features before averaging (2) normalize the new weight vector after averaging.

ronghuaiyang commented 5 years ago

thank you very much for the relay !

ronghuaiyang commented 5 years ago

hi，seasonSH another question，you said (2)normalize the new weight vector after averaging, do I need to normalize the whole weight vector or only to normalize the weight in the minibatch and update them ?