zzqzzq / bml Goto Github PK

Thank you very much for your work. I am very interested in your work and would like to quote your paper.
However, I would like to ask you about some details. I would like to know whether the test results of DeepEMD resolution from 84 to 224 that you mentioned in Table 6 were conducted directly on the pre-training model or the DeepEMD was re-trained with 224 resolution.
Thank you very much!

Thanks for your work! I think this paper is perfect. But I don't know why I can't recur the result. And the dataset, tiredImageNet, can't be used. I utilized the weight data given in the project and I get the same result with you. I'm not sure whether the parameters in the project have changed during training.

Thank you very much!