autoencoder,caeと試してきたので、次はunetを触ってみた

イメージ

f:id:programdl:20181110203911p:plain

ソース

import keras
from keras.models import load_model
from keras.datasets import cifar10
from keras.layers import Input, Dense, Conv2D, MaxPooling2D, UpSampling2D, Conv2DTranspose, concatenate
from keras.models import Model
from keras.callbacks import EarlyStopping, ModelCheckpoint
from keras.optimizers import Adam
from keras import backend as K

import os
import pickle
import numpy as np
import cv2

batch_size = 50
num_classes = 10
epochs = 100
saveDir = "./model/"
if not os.path.isdir(saveDir):
    os.makedirs(saveDir)

(x_train, y_train), (x_test, y_test) = cifar10.load_data()
x_train = x_train.astype('float32')
x_test = x_test.astype('float32')
x_train /= 255
x_test /= 255

x_val = x_test[:7000]
x_test = x_test[7000:]

def showOrigDec(orig, dec, num=10):
    import matplotlib.pyplot as plt
    n = num
    plt.figure(figsize=(20, 8))

    for i in range(n):
        # display original
        ax = plt.subplot(4, n, i+1)
        #plt.imshow(orig[i].reshape(32, 32, 3))
        plt.imshow(orig[i])
        ax.get_xaxis().set_visible(False)
        ax.get_yaxis().set_visible(False)

        # display reconstruction
        ax = plt.subplot(4, n, i +1 + n)
        #plt.imshow(dec[i].reshape(32, 32, 3))
        plt.imshow(dec[i])
        ax.get_xaxis().set_visible(False)
        ax.get_yaxis().set_visible(False)

        # display histogram
        ax = plt.subplot(4, n, i +1 + n + n)
        plt.hist((orig[i]*255).ravel(),256,[0,256])
        ax.get_xaxis().set_visible(False)
        ax.get_yaxis().set_visible(False)

        ax = plt.subplot(4, n, i +1 + n + n + n)
        color = ('b','g','r')
        for j,col in enumerate(color):
            histr = cv2.calcHist([(orig[i]*255)],[j],None,[256],[0,256])
            plt.plot(histr,color = col)
        ax.get_xaxis().set_visible(False)
        ax.get_yaxis().set_visible(False)
    plt.show()

def dice_coef(y_true, y_pred):
    y_true_f = K.flatten(y_true)
    y_pred_f = K.flatten(y_pred)
    intersection = K.sum(y_true_f * y_pred_f)
    return (2. * intersection + 1) / (K.sum(y_true_f) + K.sum(y_pred_f) + 1)

def dice_coef_loss(y_true, y_pred):
    return -dice_coef(y_true, y_pred)


def UNET():
    inputs = Input((32, 32, 3))
    conv1 = Conv2D(32, (3, 3), activation='relu', padding='same')(inputs)
    conv1 = Conv2D(32, (3, 3), activation='relu', padding='same')(conv1)
    pool1 = MaxPooling2D(pool_size=(2, 2))(conv1)
    conv2 = Conv2D(64, (3, 3), activation='relu', padding='same')(pool1)
    conv2 = Conv2D(64, (3, 3), activation='relu', padding='same')(conv2)
    pool2 = MaxPooling2D(pool_size=(2, 2))(conv2)
    conv3 = Conv2D(128, (3, 3), activation='relu', padding='same')(pool2)
    conv3 = Conv2D(128, (3, 3), activation='relu', padding='same')(conv3)
    pool3 = MaxPooling2D(pool_size=(2, 2))(conv3)
    conv4 = Conv2D(256, (3, 3), activation='relu', padding='same')(pool3)
    conv4 = Conv2D(256, (3, 3), activation='relu', padding='same')(conv4)
    pool4 = MaxPooling2D(pool_size=(2, 2))(conv4)
    conv5 = Conv2D(512, (3, 3), activation='relu', padding='same')(pool4)
    conv5 = Conv2D(512, (3, 3), activation='relu', padding='same')(conv5)

    up6 = concatenate([Conv2DTranspose(256, (2, 2), strides=(2, 2), padding='same')
                        (conv5), conv4], axis=3)
    conv6 = Conv2D(256, (3, 3), activation='relu', padding='same')(up6)
    conv6 = Conv2D(256, (3, 3), activation='relu', padding='same')(conv6)
    up7 = concatenate([Conv2DTranspose(128, (2, 2), strides=(2, 2), padding='same')
                        (conv6), conv3], axis=3)
    conv7 = Conv2D(128, (3, 3), activation='relu', padding='same')(up7)
    conv7 = Conv2D(128, (3, 3), activation='relu', padding='same')(conv7)
    up8 = concatenate([Conv2DTranspose(64, (2, 2), strides=(2, 2), padding='same')
                        (conv7), conv2], axis=3)
    conv8 = Conv2D(64, (3, 3), activation='relu', padding='same')(up8)
    conv8 = Conv2D(64, (3, 3), activation='relu', padding='same')(conv8)
    up9 = concatenate([Conv2DTranspose(32, (2, 2), strides=(2, 2), padding='same')
                        (conv8), conv1], axis=3)
    conv9 = Conv2D(32, (3, 3), activation='relu', padding='same')(up9)
    conv9 = Conv2D(32, (3, 3), activation='relu', padding='same')(conv9)
    conv10 = Conv2D(3, (3, 3), activation='sigmoid', padding='same')(conv9)
    model = Model(inputs=[inputs], outputs=[conv10])
    return model

model=UNET()
model.compile(optimizer='adam', loss=dice_coef_loss, metrics=[dice_coef])
es_cb = EarlyStopping(monitor='val_loss', patience=4, verbose=1, mode='auto')
chkpt = saveDir + 'AE_UNET.{epoch:02d}-{loss:.2f}-{val_loss:.2f}.hdf5'
cp_cb = ModelCheckpoint(filepath = chkpt, monitor='val_loss', \
                        verbose=1, save_best_only=True, mode='auto')
history = model.fit(x_train, x_train,
                    batch_size=batch_size,
                    epochs=epochs,
                    verbose=1,
                    validation_data=(x_val, x_val),
                    callbacks=[es_cb, cp_cb],
                    shuffle=True)

c10test = model.predict(x_test)
c10val = model.predict(x_val)
showOrigDec(x_test, c10test)