KaigeLi
KaigeLi
我在#16看到了相同的问题，但是没有解决我的困惑。 “Hi, the implementation of sharing the query and key branch follows Fig. 3 in the arxiv version.” 但是我在arxiv version fig. 3中看到的是key和value共享，而代码实现中是query和key共享，这有什么区别吗？
关于上采样
您好，非常感谢您的分享，我想请问您， x_seg_up = F.interpolate(x_seg, size=input.size()[2:], mode='bilinear', align_corners=True) x_seg_up = F.interpolate(x_seg_up,size=[2*i for i in input.size()[2:]], mode='bilinear', align_corners=True) 这里为什么要上采样两次，而不是直接一次上采样到原始大小呢？
I have a question, why not use object detection to achieve this goal?
How many GPUs are needed to reproduce the results?
Hi, I have tried to download the pretrained models, but I failed. So, can you provide a new download link? For example, Google Drive.
请问您实测哪个性能更好呢？ 能否提供两个分支的实验结果呢？
不知道可否分享一下该数据集？ 网上无法下载。 感激感激
Dear author, Thank you so much for such a great job! I have encountered a problem in my experiment and hope to get your guidance. In the experiment, we found...
I didn't find an implementation of mixed updating. Where is m defined and used?