试用软件

How can I reuse the activations from a layer and then merge it with another coming layer in the same neural network like in the figure?

1 次查看（过去 30 天）

显示更早的评论

Cinar Gulleroglu 2020-11-28

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/667108-how-can-i-reuse-the-activations-from-a-layer-and-then-merge-it-with-another-coming-layer-in-the-same

评论： Cinar Gulleroglu 2020-12-7

采纳的回答： Nomit Jangid

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。

采纳的回答

Nomit Jangid 2020-12-1

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/667108-how-can-i-reuse-the-activations-from-a-layer-and-then-merge-it-with-another-coming-layer-in-the-same#answer_561023

Hi Cinar,

It is not very clear in figure. if you want to add layers output you can use additionLayer function in MATLAB.

https://in.mathworks.com/help/deeplearning/ref/nnet.cnn.layer.additionlayer.html

A depth concatenation layer takes inputs that have the same height and width and concatenates them along the third dimension (the channel dimension).

https://in.mathworks.com/help/deeplearning/ref/nnet.cnn.layer.depthconcatenationlayer.html

I hope this helps!

1 个评论
显示 -1更早的评论隐藏 -1更早的评论

Cinar Gulleroglu 2020-12-7

Thank you so much, this was really helpful!

请先登录，再进行评论。

更多回答（0 个）

请先登录，再回答此问题。

类别

AI, Data Science, and Statistics Deep Learning Toolbox Image Data Workflows

在 Help Center 和 File Exchange 中查找有关 Image Data Workflows 的更多信息

标签

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Translated by

试用软件