Skip to main content
. 2022 Feb 6;22(3):1232. doi: 10.3390/s22031232

Figure 3.

Figure 3

The architecture of the proposed model: the input spectrogram is passed through the CNN part of the network to extract features, which are then passed through the LSTM part so that the long-term dependencies between them are identified and memorized. Finally, the combined features are fed into the output layer that predicts the class to which the input belongs to.