A Detection Algorithm for Audio Adversarial Examples in EI-Enhanced Automatic Speech Recognition

Table 1

Two-class neural network architecture.

Layer (type)	Output shape
dense_1 (dense)	(None, 64)
activation_1 (activation)	(None, 64)
dropout_1 (dropout)	(None, 64)
dense_2 (dense)	(None, 64)
activation_2 (activation)	(None, 64)
dropout_2 dDropout)	(None, 64)
dense_3 (dense)	(None, 1)
activation_3 (activation)	(None, 1)