search

BeierZhu / xERM

[AAAI 2022 Oral] This is a Pytorch implementation of the AAAI 2022 paper "Cross-Domain Empirical Risk Minimization for Unbiased Long-tailed Classification"
33 stars 7 forks source link

how to get a balanced model #3

Closed asd123-1 closed 2 years ago

asd123-1 commented 2 years ago

Dear Beier: I'm very interested in your work！In your code，I can get a imbalanced model by TDE, but I don't how do you get the balanced model ?

asd123-1 commented 2 years ago

I'm sorry I made a mistake. I understand that TDE is a balanced model

asd123-1 commented 2 years ago

I read your code carefully again, corresponding to your paper, and understood your code， TDE is a balanced model, TE is an imbalanced model. Is that right?

BeierZhu commented 2 years ago

You are correct. Similarly, CE is an imbalanced model and CE with Post-Compensate is a balanced model.