Table 3 CNN model summary for spectrogram classification.

From: Ensemble learning for biomedical signal classification: a high-accuracy framework using spectrograms from percussion and palpation

Layer (Type)

Output shape

Param #

Input Layer

(128, 128, 1)

0

Conv2D (32 filters, 3\(\times\)3)

(126, 126, 32)

320

MaxPooling2D (2\(\times\)2)

(63, 63, 32)

0

Conv2D (64 filters, 3\(\times\)3)

(61, 61, 64)

18,496

MaxPooling2D (2\(\times\)2)

(30, 30, 64)

0

Flatten

(57,600)

0

Dense (128)

(128)

7,372,928

Dropout (0.5)

(128)

0

Dense (8, softmax)

(8)

1,032

Total parameters

–

7,392,776

Back to article page