Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

关于ViG解码器工作的期待 #196

Open
Joazs opened this issue Apr 11, 2023 · 3 comments
Open

关于ViG解码器工作的期待 #196

Joazs opened this issue Apr 11, 2023 · 3 comments

Comments

@Joazs
Copy link

Joazs commented Apr 11, 2023

您好,我阅读论文发现。您都是将ViG作为一个主干网络后接经典CNN完成任务的。
请问像ViG这种图神经网络主干做语义分割任务,有没有其他专门关于图节点考虑的Decoder?

@iamhankai
Copy link
Member

目前还没见到端到端用GNN的,我觉得用GNN做decoder是个值得探索的课题

@Joazs
Copy link
Author

Joazs commented May 6, 2023

您好,我在阅读ViG代码的时候发现DyGraphConv2d方法下,有一个y=F.avg_pool2d(x, self.r, self.r)。然后再计算x与y的TopK距离获得邻居,虽然只在前两个下采样之前进行这种操作,但是我不知道这种操作的意义。为什么不直接计算x的TopK距离呢?
我尝试将这个y去掉,发现效果竟然不比原来好。

@iamhankai
Copy link
Member

iamhankai commented May 7, 2023

这个目的是为了降低计算量,否则节点数太多了

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants