Yunsheng Li

Which model do you mean?

I get it. But I have large numbers of models. I have to spend sometime to find the one you need. Are you in a rush?

I know the CVPR deadline is coming. I'm facing the same problem like you. I would try to find it as soon as possible. But I don't think you should...

You can download M(0) for GTA5 with the link https://drive.google.com/file/d/1j9VEVJh1eUMr7BqqaG8DfPsDxqUUES4d/view?usp=sharing. I cannot find M(0) for SYNTHIA immediately. You can train Deeplab with only the data from synthia dataset easily. If...

Looks weird to me. Are you in China. Maybe google drive is blocked.

I don't get your question. Do you mean to resume the train process? I think there is a parameter for resume. You can also add some codes to do that.

All the models are consistent with the description in the paper. Using M0 as an example, if you take a look at Table 1 in the paper, you can find...

Like you mentioned, when J=1, Dynamic Shift-Max is just Dynamic ReLU with the expression like y=max(a1*x1+b1*x1) (x1 is the first channel of the feature map, a1,b1 are the dynamic coefficients)....

Oh sorry, my writing is incorrect. The expression of Dynamic ReLU is y=max(a1x1, b1x1). It just picks up the feature point with stronger activation. For the meaning of the input...

There is no inconsistency for M0. You might read the old version of our paper.