How to quickly find the layer by its name in deep learning?

Question

cui,xingxing 2020-4-23

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/520179-how-to-quickly-find-the-layer-by-its-name-in-deep-learning

评论： cui,xingxing 2020-6-20

In the deep learning toolbox, I cannot easily manipulate a specific layer. For example, I know the name of a layer. How can I quickly find it in the layer array?

layers = [convolution2dLayer(3,3,'name','aa');
    reluLayer('name','bb');
    convolution2dLayer(3,3,'name','cc')]

For example, if I want to find a convolutional layer named "cc", can I only find one by one in a for loop? When there are many layers(There are dozens of layers), this method is very inefficient!

for i = 1:length(layers)
    if strcmp(layers(i).Name,'cc')
        myfindLayer = layers(i);
        break;
    end
end

If only

myfindLayer = layers ('cc')

can be operated like this! I hope the official will consider such a method, easy to operate!

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Ayush Laddha 2020-6-18

1
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/520179-how-to-quickly-find-the-layer-by-its-name-in-deep-learning#answer_453096

在 MATLAB Online 中打开

I understand that you wish to search a layer based upon its name and not using its index. You can make use of the ‘Name’ property of a layer. Let’s consider the layer array you mentioned in the question and you want to find the layer with name ‘cc’.

You can use –

layers({layers.Name} == "cc")

or

layers(strcmp({layers.Name}, 'cc'))