cchen-cc / SFDA-DPL

[MICCAI'21] Source-Free domain adaptive fundus image segmentation with denoised pseudo-labeling
51 stars 9 forks source link

how did you get pretrained model? #6

Closed chenslcool closed 1 year ago

chenslcool commented 1 year ago

train_source.py is for BEAL's UDA work, which utilize target domain data. I think it's not suitible for SFDA.