[PYTHON/KERAS] 컨볼루션 신경망 만들기 (손글씨) : 데이터 확장

■ 컨볼루션 신경망을 만드는 방법을 보여준다.

▶ 예제 코드 (PY)


import keras.layers as layers
import keras.layers.convolutional as convolutional
import keras.models as models
import keras.preprocessing.image as image
import numpy as np

np.random.seed(3)

print("데이터 로드를 시작합니다.")

trainImageDataGenerator = image.ImageDataGenerator(rescale = 1. / 255,
    rotation_range = 10,
    width_shift_range = 0.2,
    height_shift_range = 0.2,
    shear_range = 0.7,
    zoom_range = [0.9, 2.2],
    horizontal_flip = True,
    vertical_flip = True,
    fill_mode = "nearest")

trainDirectoryIterator = trainImageDataGenerator.flow_from_directory("handwriting_shape/train", target_size = (24, 24), batch_size = 3, class_mode = "categorical")

testImageDataGenerator = image.ImageDataGenerator(rescale = 1. / 255)

testDirectoryIterator = testImageDataGenerator.flow_from_directory("handwriting_shape/test", target_size = (24, 24), batch_size = 3, class_mode = "categorical")

print("데이터 로드를 종료합니다.")

print("모델 정의를 시작합니다.")

model = models.Sequential()

model.add(convolutional.Conv2D(32, kernel_size = (3, 3), activation = "relu", input_shape = (24, 24, 3)))
model.add(convolutional.Conv2D(64, (3, 3), activation = "relu"))
model.add(convolutional.MaxPooling2D(pool_size = (2, 2)))
model.add(layers.Flatten())
model.add(layers.Dense(128, activation = "relu"))
model.add(layers.Dense(3, activation = "softmax"))

model.compile(loss = "categorical_crossentropy", optimizer = "adam", metrics = ["accuracy"])

print("모델 정의를 종료합니다.")

print("모델 학습을 시작합니다.")

model.fit_generator(trainDirectoryIterator, steps_per_epoch = 15 * 100, epochs = 200, validation_data = testDirectoryIterator, validation_steps = 5)

print("모델 학습을 종료합니다.")

print("모델 평가를 시작합니다.")

evaluationList = model.evaluate_generator(testDirectoryIterator, steps = 5)

print("%s : %.2f%%" % (model.metrics_names[1], evaluationList[1] * 100))

print("모델 평가를 종료합니다.")

print("모델 사용을 시작합니다.")

outputNDArray = model.predict_generator(testDirectoryIterator, steps = 5)

np.set_printoptions(formatter = {"float": lambda x: "{0:0.3f}".format(x)})

classDictionary = testDirectoryIterator.class_indices

print(classDictionary)
print(outputNDArray)

print("모델 사용을 종료합니다.")

import keras.layers as layers

import keras.layers.convolutional as convolutional

import keras.models as models

import keras.preprocessing.image as image

import numpy as np

np.random.seed(3)

print("데이터 로드를 시작합니다.")

trainImageDataGenerator = image.ImageDataGenerator(rescale = 1. / 255,

rotation_range = 10,

width_shift_range = 0.2,

height_shift_range = 0.2,

shear_range = 0.7,

zoom_range = [0.9, 2.2],

horizontal_flip = True,

vertical_flip = True,

fill_mode = "nearest")

trainDirectoryIterator = trainImageDataGenerator.flow_from_directory("handwriting_shape/train", target_size = (24, 24), batch_size = 3, class_mode = "categorical")

testImageDataGenerator = image.ImageDataGenerator(rescale = 1. / 255)

testDirectoryIterator = testImageDataGenerator.flow_from_directory("handwriting_shape/test", target_size = (24, 24), batch_size = 3, class_mode = "categorical")

print("데이터 로드를 종료합니다.")

print("모델 정의를 시작합니다.")

model = models.Sequential()

model.add(convolutional.Conv2D(32, kernel_size = (3, 3), activation = "relu", input_shape = (24, 24, 3)))

model.add(convolutional.Conv2D(64, (3, 3), activation = "relu"))

model.add(convolutional.MaxPooling2D(pool_size = (2, 2)))

model.add(layers.Flatten())

model.add(layers.Dense(128, activation = "relu"))

model.add(layers.Dense(3, activation = "softmax"))

model.compile(loss = "categorical_crossentropy", optimizer = "adam", metrics = ["accuracy"])

print("모델 정의를 종료합니다.")

print("모델 학습을 시작합니다.")

model.fit_generator(trainDirectoryIterator, steps_per_epoch = 15 * 100, epochs = 200, validation_data = testDirectoryIterator, validation_steps = 5)

icodebroker

[PYTHON/KERAS] 컨볼루션 신경망 만들기 (손글씨) : 데이터 확장