Mu Hu

I think you are right but I remember having fixed it long ago... Anyway

> 作者你好，非常喜歡你這篇文章的idea，然後我現在是想extend到3D，但是我沒有太看懂在OREPA function裡的self.fre_init的作用，可以再解釋一下ma? 然後對於prior_tensor，我們能變成3D的嗎? Thank you for your interest Lee! The **self.freq_init** function aims at generating the basis corresponding to different spatial frequency patterns. This idea comes from the work...

> Hi, I'm wondering if you've run into any issues with numerical stability or know what may be the cause. > > With normal RepVGG, I get differences as high...

Thanks for your interest swaiwi! To make things simpler, convolution can be regarded as a linear mapping over feature vectors. Let's denote the weights as **w**. These weights will be...

> 您好，可以放开一下您谷歌云盘预训练权重的访问权限吗？需要申请权限才可下载，感谢！ 感谢关注！权限已更新

> How many parameters dose PENet have? I want to cite PENet in my paper, but I don't know the number of PENet's parameters. ENet ~131M and PENet ~133M, but...

> Thank you for your nice work. > Since the code is not yet open, I write down my question for your kernel design. > In the paper, given the...

> @JUGGHM could you please share your code with me ?maybe i can provide some suggestions The implemention actually refered to the CSPN naïve implemention by the author of CSPN,...

Worse results as well. But CSPN works as a refinement module.

> Worse results as well. But CSPN works as a refinement module. > > is there an open source version of CSPN? https://github.com/XinJCheng/CSPN