OCR

building_model.py

def math_symbol_and_digits_recognition(input_shape=(32, 32, 1)):
    regularizer = l2(0.01)
    model = Sequential()
    model.add(Input(shape=input_shape))
    model.add(Conv2D(32, (3, 3), strides=(1, 1), padding='same', 
                     kernel_initializer=glorot_uniform(seed=0), 
                     name='conv1', activity_regularizer=regularizer))
    model.add(Activation(activation='relu', name='act1'))
    model.add(MaxPool2D((2, 2), strides=(2, 2)))
    model.add(Conv2D(32, (3, 3), strides=(1, 1), padding='same', 
                     kernel_initializer=glorot_uniform(seed=0), 
                     name='conv2', activity_regularizer=regularizer))
    model.add(Activation(activation='relu', name='act2'))
    model.add(MaxPool2D((2, 2), strides=(2, 2)))
    model.add(Conv2D(64, (3, 3), strides=(1, 1), padding='same', 
                     kernel_initializer=glorot_uniform(seed=0), 
                     name='conv3', activity_regularizer=regularizer))
    model.add(Activation(activation='relu', name='act3'))
    model.add(MaxPool2D((2, 2), strides=(2, 2)))
    model.add(Flatten())
    model.add(Dropout(0.5))
    model.add(Dense(120, activation='relu', kernel_initializer=glorot_uniform(seed=0), name='fc1'))
    model.add(Dense(84, activation='relu', kernel_initializer=glorot_uniform(seed=0), name='fc2'))
    model.add(Dense(14, activation='softmax', kernel_initializer=glorot_uniform(seed=0), name='fc3'))
    
    optimizer = Adam()
    model.compile(loss='categorical_crossentropy', optimizer=optimizer, metrics=['accuracy'])
    return model

model = math_symbol_and_digits_recognition(input_shape=(32, 32, 1))
model.summary()