您的代码里要求输入.npy文件，应该是图片对应的mask。请问是否可以描述一下，如何生成这个.npy文件的，是否有相应的格式？
我尝试过用扣下行人的黑白图片，每一位按照a/127-1操作生成.npy文件，但生成的转移图像很奇怪。

作者你好，我在复现程中并没有发现您提供的PSPNet代码，于是导致了这部分的错误。请问PSPNet这部分怎么提取的？大致思路也可以~，呜呜呜

in your paper ,it's said that PSPNet is used for mask, but i didn't see that in your code.how do you do the mask?and does it matter?

Hi, I can't reproduce your results in Figure 8. I mean the baseline such as training on Market and directly testing on Duke. I want to know the network architecture, the optimizer, the learning rate, input size, embedding size, and so on. Hope that someone can help me.

The link you offered for MSMT17 downloading is invalid, could you please upload it again, thanks a lot.