classify
cliff0917 opened this issue · comments
Hello,
I have run your code,
but I want to load the trained model for classify after training by your work,
how can I achieve that?
You can save the best generator model weights at the epoch which gave you the best HM and ZSL accuracy.
You can save the best generator model weights at the epoch which gave you the best HM and ZSL accuracy.
Thank you for your reply,
Only the weights of generator?
How about encoder, SED, and classifier?
We extract the features from res101,
and we take the features and semantic embedding a as the input of encoder,
then we get var, mean, and generate z
taking the noise z and semantic embedding a as the input of generator,
and then we get the synthesis features x,
we put the x into SED, and get the reconstructed embedding b and hidden layer embedding h,
and we concate x, h, and b as the input of classifier, and finally get the classification result.
Is my understanding correct?
Can you explain your problem statement in detail? Where are you planning to add a classifier?
Is your problem solved?
Is your problem solved?
Yes