velkibaazi.py

def MyLSTMEEGNet(nb_classes=4, Chans=14, Samples=641, 
                dropoutRate=0.5, kernLength=64, F1=8, 
                D=4, F2=16, norm_rate=0.25, dropoutType='Dropout', 
                lstm_units=50):
    
    if dropoutType == 'SpatialDropout2D':
        dropoutType = SpatialDropout2D
    elif dropoutType == 'Dropout':
        dropoutType = Dropout
    else:
        raise ValueError('dropoutType must be one of SpatialDropout2D '
                         'or Dropout, passed as a string.')
    
    input_main = Input((Chans, Samples, 1))
    block1 = Conv2D(25, (1, 5), 
                    input_shape=(Chans, Samples, 1),
                    kernel_constraint=max_norm(2., axis=(0,1,2)))(input_main)
    block1 = BatchNormalization(epsilon=1e-05, momentum=0.9)(block1)
    block1 = Activation('elu')(block1)
    block1 = MaxPooling2D(pool_size=(1, 2), strides=(1, 2))(block1)
    block1 = dropoutType(dropoutRate)(block1)
  
    block2 = Conv2D(50, (1, 5),
                    kernel_constraint=max_norm(2., axis=(0,1,2)))(block1)
    block2 = BatchNormalization(epsilon=1e-05, momentum=0.9)(block2)
    block2 = Activation('elu')(block2)
    block2 = MaxPooling2D(pool_size=(1, 2), strides=(1, 2))(block2)
    block2 = dropoutType(dropoutRate)(block2)
    
    block3 = Conv2D(100, (1, 5),
                    kernel_constraint=max_norm(2., axis=(0,1,2)))(block2)
    block3 = BatchNormalization(epsilon=1e-05, momentum=0.9)(block3)
    block3 = Activation('elu')(block3)
    block3 = MaxPooling2D(pool_size=(1, 2), strides=(1, 2))(block3)
    block3 = dropoutType(dropoutRate)(block3)
    
    block4 = Conv2D(200, (1, 5),
                    kernel_constraint=max_norm(2., axis=(0,1,2)))(block3)
    block4 = BatchNormalization(epsilon=1e-05, momentum=0.9)(block4)
    block4 = Activation('elu')(block4)
    block4 = MaxPooling2D(pool_size=(1, 2), strides=(1, 2))(block4)
    block4 = dropoutType(dropoutRate)(block4)
    
    reshaped = Reshape((block4.shape[2], block4.shape[1] * block4.shape[3]))(block4)
    lstm = LSTM(lstm_units, return_sequences=True)(reshaped)
    
    flatten = Flatten()(lstm)
    
    dense = Dense(nb_classes, kernel_constraint=max_norm(norm_rate))(flatten)
    output = Activation('linear')(dense)
    
    return Model(inputs=input_main, outputs=output)