CNN Layers: high pass filters and truncation layer

Question

Nour Mohamed 2020-12-6

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/682273-cnn-layers-high-pass-filters-and-truncation-layer

回答： Srivardhan Gadila 2020-12-17

I would like to create two layers for my CNN design where the input image is high-pass filtered via convolution with 16 (4 × 4) high-pass DCT filters, and then a Truncation layer (Truncation Linear Unit) activation function.

These layers are not built in matlab deep learning toolbox. How can I make them?

Thank you in adavance!

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

Srivardhan Gadila 2020-12-17

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/682273-cnn-layers-high-pass-filters-and-truncation-layer#answer_578270

You can write your own Custom Layer using Deep Learning Toolbox in MATLAB. To get started you can refer to Define Custom Deep Learning Layers & Deep Learning Custom Layers.