Yunsheng Li

Thanks for you reply. I think generating pseudo labels is the key point of this paper. I can understand the high level idea but I do need to run some...

Thanks for you considerable reply. I have one more question. In the readme , the mIoU for HP-constrained is more than 29. But in the paper, it is 28.43 (table...

I find a bug when I try to evaluate the model with multiple GPUs. In the file "detectron-self-train/lib/nn/parallel/data_parallel.py" line 82 (mini_kwargs = dict([(k, v[i]) for k, v in kwargs.items()])), I...

single gpu is fine.

Thanks for your response. I can use single GPU to evaluate the results. But I still have trouble in generating pseudo labels. In the file threshold_search.py, it seems that file...

I find a problem I might have when I test the model. Could you let me know what is the test set you use to get the results in the...

So I think the result you show in the README file is based on BDD-Target-Test. But the link for the eval file shows that the val dataset you use is...

But I think different setting has different domain. In the table 4 of your paper, which validation set do you use?

I think for domain adaptation problem, the validation set should have the same domain as the target training set. Correct?

I see. I think the model you release in the github is based on the test set of bdd_peds_not_clear_any_daytime_val not bdd_peds_val which is shown in the eval file. I think...