Fig. 3From: Exploring convolutional, recurrent, and hybrid deep neural networks for speech and music detection in a large audio datasetBasic scheme of a fully connected neural network with five inputs, two outputs, and a single hidden layerBack to article page