The replication of gatys_ecker_betghe_2016 shows clear differences in the results. For this reason, I have contacted the author. He provided helpful tips for improving the results. This information on the hyperparameters is integrated with this PR. The most important changes are:
the use of the activations after the ReLU function.
the use of layer_weights depending on the number of channels.
random initialisation of the image and more optimisation steps.
Due to the random initalisation, an exact replication is no longer possible, but the results show a clear improvement and a more similar stylisation to the original images.
The replication of gatys_ecker_betghe_2016 shows clear differences in the results. For this reason, I have contacted the author. He provided helpful tips for improving the results. This information on the hyperparameters is integrated with this PR. The most important changes are:
layer_weights
depending on the number of channels.Due to the random initalisation, an exact replication is no longer possible, but the results show a clear improvement and a more similar stylisation to the original images.