search

ThomasWestfechtel / SKD

Combining inherent knowledge of vision-language models with unsupervised domain adaptation through self-knowledge distillation
6 stars 0 forks source link

Results on DomainNet. #2

Open Wenlve-Zhou opened 6 months ago

Wenlve-Zhou commented 6 months ago

Thanks for your great work! Could you please prove a training script and zero-shot prediction configs on DomainNet? I can't reproduce the results presented in your paper. Hope for your response~

ThomasWestfechtel commented 3 months ago

Sorry for the late reply Please have a look at the updated version: https://github.com/ThomasWestfechtel/SWG It includes the script for generating the zero-shot predictions