스터디노트 (Tensorflow 7)

ImageDataGenerator

import os
from glob import glob

import tensorflow as tf
import numpy as np
from PIL import Image

import matplotlib.pyplot as plt
%matplotlib inline

ImageDataGenerator

• 데이터를 불러오는 동시에 여러가지 전처리를 쉽게 구현 할 수 있는 tf.keras의 기능!

from tensorflow.keras.preprocessing.image import ImageDataGenerator

datagen = ImageDataGenerator(
    rotation_range=20, # 회전하는 각도를 20도 내에서 랜덤하게 회전
    width_shift_range=0.2, # 가로로 이동하는데 20% 만큼 랜덤으로 생성
    height_shift_range=0.2, # 높이로 이동하는데 20% 만큼 랜덤으로 생성
    horizontal_flip = True # 가로축 랜덤 설정
)

datagen.flow()

Data Aumentation -> Overfitting 방지

flow

• 데이터를 모두 메모리에 불러두고 사용 할 때

class Cifar10DataLoader():
    def __init__(self):
        # data load
        (self.train_x, self.train_y), \
            (self.test_x, self.test_y) = tf.keras.datasets.cifar10.load_data()
        self.input_shape = self.train_x.shape[1:]

    def scale(self, x):

        return (x / 255.0).astype(np.float32)

    def preprocess_dataset(self, dataset):

        (feature, target) = dataset

        # scaling #
        scaled_x = np.array([self.scale(x) for x in feature])

        # label encoding #
        ohe_y = np.array([tf.keras.utils.to_categorical(
            y, num_classes=10) for y in target])
        
        return scaled_x, ohe_y.squeeze(1)

    def get_train_dataset(self):
        return self.preprocess_dataset((self.train_x, self.train_y))

    def get_test_dataset(self):
        return self.preprocess_dataset((self.test_x, self.test_y))

cifar10_loader = Cifar10DataLoader()
train_x, train_y = cifar10_loader.get_train_dataset()

print(train_x.shape, train_x.dtype)
print(train_y.shape, train_y.dtype)

result = next(iter(datagen.flow((train_x, train_y))))

x, y = result

x.shape

y.shape

x[0].shape

plt.imshow(x[0])
plt.show()

flow_from_directory

train_dir = "../../datasets/mnist_png/training"

input_shape = (28, 28, 1)
batch_size = 32

gen = datagen.flow_from_directory(
    train_dir,
    target_size=input_shape[:2], # 이미지를 불러 왔을 때 어떤 크기로 줄여줄까
    batch_size= batch_size,
    color_mode = 'grayscale'
)

x, y = next(iter(gen))

x.shape, y.shape

flow_from_DataFrame

import pandas as pd

os.listdir('../../datasets/cifar/train_dataset.csv')

train_data = pd.read_csv('../../datasets/cifar/train_dataset.csv')

gen = datagen.flow_from_dataframe(
    train_data,
    x_col='path',
    y_col = 'class_name',
    target_size=(32, 32),
    color_mode='rgb',
    class_mode='categorical',
    batch_size=32
)

x, y = next(iter(gen))

x.shape

y.shape

학습

from tensorflow.keras.layers import Input, Conv2D, MaxPool2D, Flatten, Dense, Add

def build_resnet(input_shape):
    inputs = Input(input_shape)

    net = Conv2D(32, kernel_size=3, strides=2,
                 padding='same', activation='relu')(inputs)
    net = MaxPool2D()(net)
    
    net1 = Conv2D(64, kernel_size=1, padding='same', activation='relu')(net)
    net2 = Conv2D(64, kernel_size=3, padding='same', activation='relu')(net1)
    net3 = Conv2D(64, kernel_size=1, padding='same', activation='relu')(net2)
    
    net1_1 = Conv2D(64, kernel_size=1, padding='same')(net)
    net = Add()([net1_1, net3])
    
    net1 = Conv2D(64, kernel_size=1, padding='same', activation='relu')(net)
    net2 = Conv2D(64, kernel_size=3, padding='same', activation='relu')(net1)
    net3 = Conv2D(64, kernel_size=1, padding='same', activation='relu')(net2)
    
    net = Add()([net, net3])
    
    net = MaxPool2D()(net)
    
    net = Flatten()(net)
    net = Dense(10, activation="softmax")(net)

    model = tf.keras.Model(inputs=inputs, outputs=net, name='resnet')
    
    return model

model = build_resnet((32, 32, 3))
model.summary()

learning_rate = 0.03
opt = tf.keras.optimizers.Adam(learning_rate)
loss = tf.keras.losses.categorical_crossentropy

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

model.fit(gen)

어렵다...
💻 출처 : 제로베이스 데이터 취업 스쿨