emotion_detection.py

model1=keras.models.Sequential()

# Block-1
model1.add(keras.layers.Conv2D(filters=32, kernel_size=(3,3), padding='same', 
                     kernel_initializer='he_normal',
                     activation="elu", 
                     input_shape=(48,48,1)))
model1.add(keras.layers.BatchNormalization())

model1.add(keras.layers.Conv2D(filters=32, kernel_size=(3,3), padding='same', 
                     kernel_initializer='he_normal', 
                     activation="elu"))

model1.add(keras.layers.BatchNormalization())
model1.add(keras.layers.MaxPooling2D(pool_size=(2,2)))
model1.add(keras.layers.Dropout(0.2))

# Block-2
model1.add(keras.layers.Conv2D(filters=64, kernel_size=(3,3), padding='same', 
                     kernel_initializer='he_normal',
                     activation="elu"))

model1.add(keras.layers.BatchNormalization())

model1.add(keras.layers.Conv2D(filters=64,kernel_size=(3,3),padding='same',
                     kernel_initializer='he_normal', 
                     activation="elu"))

model1.add(keras.layers.BatchNormalization())
model1.add(keras.layers.MaxPooling2D(pool_size=(2,2)))
model1.add(keras.layers.Dropout(0.2))

# Block-3
model1.add(keras.layers.Conv2D(filters=128, kernel_size=(3,3), padding='same', 
                     kernel_initializer='he_normal', 
                     activation="elu"))
model1.add(keras.layers.BatchNormalization())
model1.add(keras.layers.Conv2D(filters=128, kernel_size=(3,3), padding='same', 
                     kernel_initializer='he_normal',
                     activation="elu"))

model1.add(keras.layers.BatchNormalization())
model1.add(keras.layers.MaxPooling2D(pool_size=(2,2)))
model1.add(keras.layers.Dropout(0.2))

# Block-4
model1.add(keras.layers.Flatten())
model1.add(keras.layers.Dense(64, activation="elu", kernel_initializer='he_normal'))
model1.add(keras.layers.BatchNormalization())
model1.add(keras.layers.Dropout(0.5))

# Block-5
model1.add(keras.layers.Dense(7, activation="softmax", kernel_initializer='he_normal'))

print(model1.summary())