Fig. 2

Architecture of a 1-dimensional convolutional neural network (1D-CNN). In this example, a STP is presented to the input layer. The trained 1D-CNN computes the probabilities for NSTP and STP, respectively