Table 1.

Proposed Network Configuration.

Texture Classification Analysis
Layer name	Layer description
Input	9 Mini-batch, 10 Sequence, 1 Channel values
RNN (LSTM Layer 1)	Input = 1, Output = 256, Activation = tanh
RNN (LSTM Layer 2)	Input = 256, Output = 256, Activation = tanh, Dropout = 0.7
FC layer 1	Input = 512, Output = 512, Activation = Leaky_relu (alpha 0.2), Dropout = 0.7
FC layer 2	Input = 512, Output = 1024, Activation = Leaky_relu (alpha 0.2)
Softmax	Input = 1024, Output = 9