Figure 8 | Scientific Reports

Figure 8

From: Unsupervised anomaly detection for earthquake detection on Korea high-speed trains using autoencoder-based deep learning models

Figure 8

Autoencoder architecture. An autoencoder comprises an encoder and a decoder, with an intermediary latent vector capturing essential features of the input data. This latent vector serves as input to the decoder, generating output data that encapsulates the key features of the input. In certain instances, the encoder and decoder components of this autoencoder are adapted as Conv-AE or LSTM-AE by substituting them with CNN or LSTM.

Back to article page