Visualizing different layers of neural network

[Sohrab_Salimian]

Hello,

I was just wondering whether there was a specific method for visualizing the layers in cnn, i want to visualize from the first conv1, conv2 etc layers all the way to the fully connected layers,

Also my network contains a series of nn.sequential components is there any way to visualize the components that make that up as well?

[jytug (Filip Binkiewicz)]

You should take a look at this response from Francisco Massa:

How to extract features of an image from a trained model

To complement @apaszke reply, once you have a trained model, if you want to extract the result of an intermediate layer (say fc7 after the relu), you have a couple of possibilities. You can either reconstruct the classifier once the model was instantiated, as in the following example: import torch import torch.nn as nn from torchvision import models model = models.alexnet(pretrained=True) # remove last fully-connected layer new_classifier = nn.Sequential(*list(model.classifier.children())[:-…

Having extracted the intermediate outputs of a convolutional network, you can visualise them using matplotlib