FaPN-full
FaPN-full copied to clipboard
[ICCV 2021] FaPN: Feature-aligned Pyramid Network for Dense Image Prediction
没有找到fan.py中的 detectron2.layer文件
您好,您的工作是一份非常有价值的工作。由于我在pytorch1.7.1下,编译不了您提供的可变性卷积的库。所以我尝试用mmd实现您的代码,我遇到了一个问题: self.dcpack_L2 = dcn_v2(out_nc, out_nc, 3, stride=1, padding=1, dilation=1, deformable_groups=8, extra_offset_mask=True) 这行代码的mask是根据x算的,还是根据offset算的?您知道具体的计算方法吗?
When running the network, I encountered this problem. Through debugging, I found that the offset in DCN's forword function is a type of float16.So I think this might be the...
After watching your code, i have a question, about every layer after "FSM" and "FAM" model the output channels are same.Do I understand that right? And i want to know...
FSM和FAM
您好,看了您的论文,希望借助一下您论文中的FSM和FAM模块,直接插入到我的网络中,但我没有找到这两个模块的单独实现,请问这两个模块的单独实现在哪里呢,谢谢