The output of your convolutional layer is generally passed through the ReLU activation functionality to bring non-linearity on the model. It will require the feature map and replaces each of the destructive values with zero. Zero-padding—It allows us to control the spatial size of the output volume by padding https://financefeeds.com/cardano-long-term-history-shows-how-1fuels-advanced-identical-tech-is-the-best-pick-for-january/