Question about the Matlab Wasserstein GAN example

4 次查看（过去 30 天）
Mohammed
Mohammed 2024-5-21
编辑： Malay Agarwal 2024-5-22
The original Wasserstein gan paper suggest removing the Critic's last dense layer activation function(sigmoid) such that the output value is not limited to fake or real. The posted example still uses sigmoid layer, am I right?

请先登录，再回答此问题。

回答（1 个）

Malay Agarwal
Malay Agarwal 2024-5-22
编辑：Malay Agarwal 2024-5-22
Hi @Mohammed,
The diagram of the Discriminator model in the example (https://www.mathworks.com/help/deeplearning/ug/trainwasserstein-gan-with-gradient-penalty-wgan-gp.html) shows that the model does have a “sigmoid” layer at the end:
This can also be confirmed by looking at how the Discriminator model is defined:
layersD = [
imageInputLayer(inputSize,Normalization="none")
convolution2dLayer(filterSize,numFilters,Stride=2,Padding="same")
leakyReluLayer(scale)
convolution2dLayer(filterSize,2*numFilters,Stride=2,Padding="same")
layerNormalizationLayer
leakyReluLayer(scale)
convolution2dLayer(filterSize,4*numFilters,Stride=2,Padding="same")
layerNormalizationLayer
leakyReluLayer(scale)
convolution2dLayer(filterSize,8*numFilters,Stride=2,Padding="same")
layerNormalizationLayer
leakyReluLayer(scale)
convolution2dLayer(4,1)
sigmoidLayer]; % Notice the sigmoid layer at the end
Hope this helps!

类别

在 Help Center 和 File Exchange 中查找有关 Statistics and Machine Learning Toolbox 的更多信息

标签

Community Treasure Hunt

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

Start Hunting!

Translated by