Convolution layer (CONV) The convolution layer (CONV) uses filters that perform convolution operations as it really is scanning the input $I$ with regard to its dimensions. Its hyperparameters contain the filter size $F$ and stride $S$. The resulting output $O$ is called characteristic map or activation map. The LeNet https://financefeeds.com/foresight-ventures-launches-latest-payfi-report-new-frontier-of-blockchain-payment-innovation/