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

最优频率分量问题 #19

Open
Lewis0427 opened this issue Sep 6, 2021 · 4 comments
Open

最优频率分量问题 #19

Lewis0427 opened this issue Sep 6, 2021 · 4 comments

Comments

@Lewis0427
Copy link

请问top里面的最优频率分量 是不是只适用于77,一旦改变尺寸为4848 最优频率分量是否需要更换,还是仍起作用?请指点一下

@cfzd
Copy link
Owner

cfzd commented Sep 7, 2021

@Lewis0427
我们目前的实验结果表明不需要更换,我们在COCO数据集(图像大小会变化)的网络也是一样使用7x7下的频率分量。可以参考一下这部分代码:

FcaNet/model/layer.py

Lines 53 to 59 in aa5fb63

x_pooled = x
if h != self.dct_h or w != self.dct_w:
x_pooled = torch.nn.functional.adaptive_avg_pool2d(x, (self.dct_h, self.dct_w))
# If you have concerns about one-line-change, don't worry. :)
# In the ImageNet models, this line will never be triggered.
# This is for compatibility in instance segmentation and object detection.
y = self.dct_layer(x_pooled)

@Lewis0427
Copy link
Author

好的知道了 谢谢您

@Lewis0427
Copy link
Author

MultiSpectralDCTLayer类当中选择learnable random init,learnable DCT init和fixed DCT init,fixed random init会影响不同任务的效果吗 ?应该怎么选择呢?

@cfzd
Copy link
Owner

cfzd commented Sep 8, 2021

@Lewis0427
如果你不是做实验探究这几个的区别的话,应该始终选择fixed DCT init.

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