Pass some images through a pretrained CNN and visualize the feature maps from an intermediate layer (e.g., after the first or second convolutional layer).
This helps you see the types of features that CNNs pick up on early in the network and provides intuition on what kinds of patterns your model may learn.
Pass some images through a pretrained CNN and visualize the feature maps from an intermediate layer (e.g., after the first or second convolutional layer). This helps you see the types of features that CNNs pick up on early in the network and provides intuition on what kinds of patterns your model may learn.