python - 如何将图像分类为噪声图像和无噪声图像？

Question

我有包含噪声和无噪声图像的数据集。噪声图像是使用不同的 ISO 和亮度条件生成的。我尝试使用 VGG16 预训练的 CNN 模型（使用 Python 和 Keras 库）对它们进行分类，但它无法对新的图像集进行分类。所以我想知道如何解决这个问题？我们可以通过使用 VGG16、InceptionV3 等 CNN 模型来解决这个问题吗？有没有其他方法可以解决？

VGG16 CNN 模型的代码：

#XX: total dataset containing 1000 images (500: Noisy, 500:Noise Free) 
# Original Dimension of image:(4000, 3000, 3)
#YY: Contains class for each image in dataset
import numpy as np
XX=np.array(XX).reshape(-1,224,224,3)
import pandas as pd
YY = pd.get_dummies(YY)
from sklearn.model_selection import train_test_split
x_train,x_test,y_train,y_test=train_test_split(XX,YY,test_size=0.3,stratify=YY)

#Model Creation
from keras.models import Model
from keras.layers import Flatten, Dense,Dropout
from keras.applications import VGG16
from keras.optimizers import SGD
num_classes=2
IMAGE_SIZE = [224,224]  

vgg = VGG16(input_shape = IMAGE_SIZE + [3], weights = 'imagenet', include_top = False)
#For freezing the first 8 layers
for layer in vgg.layers[:8]:
       vgg.trainable = False

x = Flatten()(vgg.output)
x=Dropout(0.2)(x)
x = Dense(num_classes, activation = 'softmax')(x)  
model = Model(inputs = vgg.input, outputs = x)
opt = SGD(lr=0.000001)
model.compile(loss = "categorical_crossentropy", optimizer = opt, metrics=['accuracy'])

#ImageDataGenerator
from keras.preprocessing.image import ImageDataGenerator
BATCH_SIZE = 32
def create_datagen():
    return ImageDataGenerator(
        zoom_range=0.15,  
        fill_mode='constant',
        cval=0.,  
        horizontal_flip=True,  
        vertical_flip=True, 
        rotation_range=360
    )

data_generator = create_datagen().flow(x_train, y_train, batch_size=BATCH_SIZE)

#Fit model
history = model.fit_generator(
    data_generator,
    steps_per_epoch=x_train.shape[0] / BATCH_SIZE,
    epochs=300,
    validation_data=(x_test, y_test) 
)

得到以下结果： 在 122 Epoch 之后添加结果

Epoch 123/300
22/21 [==============================] - 377s 17s/step - loss: 0.3379 - accuracy: 0.9372 - val_loss: 0.1180 - val_accuracy: 0.9734
Epoch 124/300
22/21 [==============================] - 376s 17s/step - loss: 0.3511 - accuracy: 0.9372 - val_loss: 0.1144 - val_accuracy: 0.9734
Epoch 125/300
22/21 [==============================] - 377s 17s/step - loss: 0.2772 - accuracy: 0.9415 - val_loss: 0.1156 - val_accuracy: 0.9734
Epoch 126/300
22/21 [==============================] - 376s 17s/step - loss: 0.4165 - accuracy: 0.9173 - val_loss: 0.1126 - val_accuracy: 0.9734
Epoch 127/300
22/21 [==============================] - 377s 17s/step - loss: 0.3607 - accuracy: 0.9244 - val_loss: 0.1108 - val_accuracy: 0.9734
Epoch 128/300
22/21 [==============================] - 376s 17s/step - loss: 0.4563 - accuracy: 0.9230 - val_loss: 0.1123 - val_accuracy: 0.9734
Epoch 129/300
22/21 [==============================] - 376s 17s/step - loss: 0.4268 - accuracy: 0.9230 - val_loss: 0.1071 - val_accuracy: 0.9767
Epoch 130/300
22/21 [==============================] - 376s 17s/step - loss: 0.3658 - accuracy: 0.9344 - val_loss: 0.1059 - val_accuracy: 0.9734
Epoch 131/300
22/21 [==============================] - 376s 17s/step - loss: 0.4448 - accuracy: 0.9187 - val_loss: 0.1078 - val_accuracy: 0.9734
Epoch 132/300
22/21 [==============================] - 377s 17s/step - loss: 0.5164 - accuracy: 0.9201 - val_loss: 0.1059 - val_accuracy: 0.9767
Epoch 133/300
22/21 [==============================] - 378s 17s/step - loss: 0.4034 - accuracy: 0.9372 - val_loss: 0.1040 - val_accuracy: 0.9767
Epoch 134/300
22/21 [==============================] - 377s 17s/step - loss: 0.4147 - accuracy: 0.9230 - val_loss: 0.1019 - val_accuracy: 0.9767
Epoch 135/300
22/21 [==============================] - 376s 17s/step - loss: 0.4059 - accuracy: 0.9330 - val_loss: 0.0975 - val_accuracy: 0.9801
Epoch 136/300
22/21 [==============================] - 376s 17s/step - loss: 0.3722 - accuracy: 0.9230 - val_loss: 0.0961 - val_accuracy: 0.9801
Epoch 137/300
22/21 [==============================] - 377s 17s/step - loss: 0.3892 - accuracy: 0.9173 - val_loss: 0.0976 - val_accuracy: 0.9767
Epoch 138/300
22/21 [==============================] - 377s 17s/step - loss: 0.3381 - accuracy: 0.9415 - val_loss: 0.0978 - val_accuracy: 0.9767
Epoch 139/300
22/21 [==============================] - 376s 17s/step - loss: 0.2899 - accuracy: 0.9401 - val_loss: 0.0967 - val_accuracy: 0.9767
Epoch 140/300
22/21 [==============================] - 377s 17s/step - loss: 0.4291 - accuracy: 0.9230 - val_loss: 0.0927 - val_accuracy: 0.9801
Epoch 141/300
22/21 [==============================] - 378s 17s/step - loss: 0.2823 - accuracy: 0.9330 - val_loss: 0.0896 - val_accuracy: 0.9801
Epoch 142/300
22/21 [==============================] - 377s 17s/step - loss: 0.1845 - accuracy: 0.9501 - val_loss: 0.0878 - val_accuracy: 0.9801
Epoch 143/300
22/21 [==============================] - 377s 17s/step - loss: 0.3070 - accuracy: 0.9244 - val_loss: 0.0880 - val_accuracy: 0.9801
Epoch 144/300
22/21 [==============================] - 377s 17s/step - loss: 0.4041 - accuracy: 0.9258 - val_loss: 0.0836 - val_accuracy: 0.9801
Epoch 145/300
22/21 [==============================] - 377s 17s/step - loss: 0.2355 - accuracy: 0.9330 - val_loss: 0.0812 - val_accuracy: 0.9801
Epoch 146/300
22/21 [==============================] - 377s 17s/step - loss: 0.3423 - accuracy: 0.9372 - val_loss: 0.0811 - val_accuracy: 0.9801
Epoch 147/300
22/21 [==============================] - 376s 17s/step - loss: 0.2270 - accuracy: 0.9415 - val_loss: 0.0821 - val_accuracy: 0.9801
Epoch 148/300
22/21 [==============================] - 378s 17s/step - loss: 0.3257 - accuracy: 0.9287 - val_loss: 0.0807 - val_accuracy: 0.9801
Epoch 149/300
22/21 [==============================] - 377s 17s/step - loss: 0.3776 - accuracy: 0.9401 - val_loss: 0.0820 - val_accuracy: 0.9801
Epoch 150/300
22/21 [==============================] - 377s 17s/step - loss: 0.2924 - accuracy: 0.9515 - val_loss: 0.0796 - val_accuracy: 0.9834
Epoch 151/300
22/21 [==============================] - 377s 17s/step - loss: 0.3149 - accuracy: 0.9387 - val_loss: 0.0789 - val_accuracy: 0.9834
Epoch 152/300
22/21 [==============================] - 377s 17s/step - loss: 0.2694 - accuracy: 0.9372 - val_loss: 0.0803 - val_accuracy: 0.9834
Epoch 153/300
22/21 [==============================] - 376s 17s/step - loss: 0.3450 - accuracy: 0.9344 - val_loss: 0.0789 - val_accuracy: 0.9834
Epoch 154/300
22/21 [==============================] - 376s 17s/step - loss: 0.2978 - accuracy: 0.9401 - val_loss: 0.0767 - val_accuracy: 0.9834
Epoch 155/300
22/21 [==============================] - 377s 17s/step - loss: 0.3924 - accuracy: 0.9415 - val_loss: 0.0759 - val_accuracy: 0.9834
Epoch 156/300
22/21 [==============================] - 377s 17s/step - loss: 0.2532 - accuracy: 0.9358 - val_loss: 0.0766 - val_accuracy: 0.9834
Epoch 157/300
22/21 [==============================] - 377s 17s/step - loss: 0.3752 - accuracy: 0.9315 - val_loss: 0.0722 - val_accuracy: 0.9867
Epoch 158/300
22/21 [==============================] - 377s 17s/step - loss: 0.3317 - accuracy: 0.9401 - val_loss: 0.0740 - val_accuracy: 0.9867
Epoch 159/300
22/21 [==============================] - 376s 17s/step - loss: 0.3688 - accuracy: 0.9344 - val_loss: 0.0737 - val_accuracy: 0.9834
Epoch 160/300
22/21 [==============================] - 376s 17s/step - loss: 0.3354 - accuracy: 0.9401 - val_loss: 0.0732 - val_accuracy: 0.9834
Epoch 161/300
22/21 [==============================] - 376s 17s/step - loss: 0.2827 - accuracy: 0.9372 - val_loss: 0.0731 - val_accuracy: 0.9834
Epoch 162/300
22/21 [==============================] - 378s 17s/step - loss: 0.2141 - accuracy: 0.9458 - val_loss: 0.0721 - val_accuracy: 0.9834
Epoch 163/300
22/21 [==============================] - 378s 17s/step - loss: 0.3395 - accuracy: 0.9287 - val_loss: 0.0694 - val_accuracy: 0.9867
Epoch 164/300
22/21 [==============================] - 376s 17s/step - loss: 0.2528 - accuracy: 0.9472 - val_loss: 0.0692 - val_accuracy: 0.9834
Epoch 165/300
22/21 [==============================] - 376s 17s/step - loss: 0.2318 - accuracy: 0.9501 - val_loss: 0.0691 - val_accuracy: 0.9834
Epoch 166/300
22/21 [==============================] - 376s 17s/step - loss: 0.2182 - accuracy: 0.9486 - val_loss: 0.0676 - val_accuracy: 0.9867
Epoch 167/300
22/21 [==============================] - 378s 17s/step - loss: 0.3260 - accuracy: 0.9387 - val_loss: 0.0665 - val_accuracy: 0.9867
Epoch 168/300
22/21 [==============================] - 376s 17s/step - loss: 0.3035 - accuracy: 0.9415 - val_loss: 0.0671 - val_accuracy: 0.9867
Epoch 169/300
22/21 [==============================] - 377s 17s/step - loss: 0.2598 - accuracy: 0.9458 - val_loss: 0.0657 - val_accuracy: 0.9867
Epoch 170/300
22/21 [==============================] - 377s 17s/step - loss: 0.2582 - accuracy: 0.9529 - val_loss: 0.0658 - val_accuracy: 0.9867
Epoch 171/300
22/21 [==============================] - 377s 17s/step - loss: 0.3038 - accuracy: 0.9529 - val_loss: 0.0689 - val_accuracy: 0.9867
Epoch 172/300
22/21 [==============================] - 377s 17s/step - loss: 0.2628 - accuracy: 0.9486 - val_loss: 0.0679 - val_accuracy: 0.9867
Epoch 173/300
22/21 [==============================] - 377s 17s/step - loss: 0.2119 - accuracy: 0.9558 - val_loss: 0.0687 - val_accuracy: 0.9867
Epoch 174/300
22/21 [==============================] - 377s 17s/step - loss: 0.3272 - accuracy: 0.9358 - val_loss: 0.0702 - val_accuracy: 0.9867
Epoch 175/300
22/21 [==============================] - 377s 17s/step - loss: 0.2615 - accuracy: 0.9358 - val_loss: 0.0700 - val_accuracy: 0.9867
Epoch 176/300
22/21 [==============================] - 376s 17s/step - loss: 0.3187 - accuracy: 0.9444 - val_loss: 0.0695 - val_accuracy: 0.9867
Epoch 177/300
22/21 [==============================] - 378s 17s/step - loss: 0.2071 - accuracy: 0.9515 - val_loss: 0.0679 - val_accuracy: 0.9867
Epoch 178/300
22/21 [==============================] - 377s 17s/step - loss: 0.3139 - accuracy: 0.9401 - val_loss: 0.0661 - val_accuracy: 0.9867
Epoch 179/300
22/21 [==============================] - 377s 17s/step - loss: 0.2771 - accuracy: 0.9429 - val_loss: 0.0643 - val_accuracy: 0.9867
Epoch 180/300
22/21 [==============================] - 377s 17s/step - loss: 0.2328 - accuracy: 0.9515 - val_loss: 0.0640 - val_accuracy: 0.9867
Epoch 181/300
22/21 [==============================] - 376s 17s/step - loss: 0.1790 - accuracy: 0.9586 - val_loss: 0.0624 - val_accuracy: 0.9867
Epoch 182/300
22/21 [==============================] - 377s 17s/step - loss: 0.3432 - accuracy: 0.9515 - val_loss: 0.0614 - val_accuracy: 0.9867
Epoch 183/300
22/21 [==============================] - 377s 17s/step - loss: 0.2550 - accuracy: 0.9444 - val_loss: 0.0638 - val_accuracy: 0.9867
Epoch 184/300
22/21 [==============================] - 377s 17s/step - loss: 0.2734 - accuracy: 0.9429 - val_loss: 0.0641 - val_accuracy: 0.9867
Epoch 185/300
22/21 [==============================] - 377s 17s/step - loss: 0.2693 - accuracy: 0.9458 - val_loss: 0.0620 - val_accuracy: 0.9867
Epoch 186/300
22/21 [==============================] - 378s 17s/step - loss: 0.2916 - accuracy: 0.9387 - val_loss: 0.0626 - val_accuracy: 0.9867
Epoch 187/300
22/21 [==============================] - 377s 17s/step - loss: 0.2710 - accuracy: 0.9572 - val_loss: 0.0614 - val_accuracy: 0.9867
Epoch 188/300
22/21 [==============================] - 377s 17s/step - loss: 0.2273 - accuracy: 0.9544 - val_loss: 0.0592 - val_accuracy: 0.9867
Epoch 189/300
22/21 [==============================] - 377s 17s/step - loss: 0.2588 - accuracy: 0.9501 - val_loss: 0.0579 - val_accuracy: 0.9867
Epoch 190/300
22/21 [==============================] - 376s 17s/step - loss: 0.2744 - accuracy: 0.9501 - val_loss: 0.0560 - val_accuracy: 0.9867
Epoch 191/300
22/21 [==============================] - 378s 17s/step - loss: 0.1692 - accuracy: 0.9586 - val_loss: 0.0566 - val_accuracy: 0.9867
Epoch 192/300
22/21 [==============================] - 377s 17s/step - loss: 0.1953 - accuracy: 0.9486 - val_loss: 0.0561 - val_accuracy: 0.9867
Epoch 193/300
22/21 [==============================] - 377s 17s/step - loss: 0.2281 - accuracy: 0.9544 - val_loss: 0.0550 - val_accuracy: 0.9867
Epoch 194/300
22/21 [==============================] - 378s 17s/step - loss: 0.1234 - accuracy: 0.9586 - val_loss: 0.0541 - val_accuracy: 0.9867
Epoch 195/300
22/21 [==============================] - 376s 17s/step - loss: 0.2385 - accuracy: 0.9444 - val_loss: 0.0540 - val_accuracy: 0.9867
Epoch 196/300
22/21 [==============================] - 377s 17s/step - loss: 0.2519 - accuracy: 0.9444 - val_loss: 0.0548 - val_accuracy: 0.9867
Epoch 197/300
22/21 [==============================] - 376s 17s/step - loss: 0.2070 - accuracy: 0.9572 - val_loss: 0.0533 - val_accuracy: 0.9867
Epoch 198/300
22/21 [==============================] - 378s 17s/step - loss: 0.2044 - accuracy: 0.9572 - val_loss: 0.0530 - val_accuracy: 0.9867
Epoch 199/300
22/21 [==============================] - 377s 17s/step - loss: 0.2041 - accuracy: 0.9558 - val_loss: 0.0533 - val_accuracy: 0.9867
Epoch 200/300
22/21 [==============================] - 377s 17s/step - loss: 0.3268 - accuracy: 0.9315 - val_loss: 0.0522 - val_accuracy: 0.9867
Epoch 201/300
22/21 [==============================] - 377s 17s/step - loss: 0.1605 - accuracy: 0.9601 - val_loss: 0.0514 - val_accuracy: 0.9867
Epoch 202/300
22/21 [==============================] - 376s 17s/step - loss: 0.2490 - accuracy: 0.9486 - val_loss: 0.0505 - val_accuracy: 0.9867
Epoch 203/300
22/21 [==============================] - 376s 17s/step - loss: 0.2290 - accuracy: 0.9544 - val_loss: 0.0510 - val_accuracy: 0.9867
Epoch 204/300
22/21 [==============================] - 377s 17s/step - loss: 0.2597 - accuracy: 0.9429 - val_loss: 0.0503 - val_accuracy: 0.9867
Epoch 205/300
22/21 [==============================] - 377s 17s/step - loss: 0.1400 - accuracy: 0.9643 - val_loss: 0.0505 - val_accuracy: 0.9867
Epoch 206/300
22/21 [==============================] - 377s 17s/step - loss: 0.2330 - accuracy: 0.9558 - val_loss: 0.0492 - val_accuracy: 0.9867
Epoch 207/300
22/21 [==============================] - 376s 17s/step - loss: 0.2049 - accuracy: 0.9586 - val_loss: 0.0492 - val_accuracy: 0.9867
Epoch 208/300
22/21 [==============================] - 377s 17s/step - loss: 0.2605 - accuracy: 0.9458 - val_loss: 0.0464 - val_accuracy: 0.9867
Epoch 209/300
22/21 [==============================] - 378s 17s/step - loss: 0.1611 - accuracy: 0.9658 - val_loss: 0.0471 - val_accuracy: 0.9867
Epoch 210/300
22/21 [==============================] - 378s 17s/step - loss: 0.2058 - accuracy: 0.9501 - val_loss: 0.0462 - val_accuracy: 0.9867
Epoch 211/300
22/21 [==============================] - 378s 17s/step - loss: 0.2022 - accuracy: 0.9515 - val_loss: 0.0455 - val_accuracy: 0.9867
Epoch 212/300
22/21 [==============================] - 377s 17s/step - loss: 0.1029 - accuracy: 0.9729 - val_loss: 0.0451 - val_accuracy: 0.9867
Epoch 213/300
22/21 [==============================] - 377s 17s/step - loss: 0.1800 - accuracy: 0.9472 - val_loss: 0.0453 - val_accuracy: 0.9867
Epoch 214/300
22/21 [==============================] - 377s 17s/step - loss: 0.1471 - accuracy: 0.9700 - val_loss: 0.0446 - val_accuracy: 0.9867
Epoch 215/300
22/21 [==============================] - 401s 18s/step - loss: 0.2121 - accuracy: 0.9686 - val_loss: 0.0453 - val_accuracy: 0.9867
Epoch 216/300
22/21 [==============================] - 402s 18s/step - loss: 0.1735 - accuracy: 0.9586 - val_loss: 0.0444 - val_accuracy: 0.9867
Epoch 217/300
22/21 [==============================] - 402s 18s/step - loss: 0.1517 - accuracy: 0.9486 - val_loss: 0.0447 - val_accuracy: 0.9867
Epoch 218/300
22/21 [==============================] - 401s 18s/step - loss: 0.1743 - accuracy: 0.9529 - val_loss: 0.0443 - val_accuracy: 0.9867
Epoch 219/300
22/21 [==============================] - 402s 18s/step - loss: 0.2412 - accuracy: 0.9586 - val_loss: 0.0440 - val_accuracy: 0.9867
Epoch 220/300
22/21 [==============================] - 399s 18s/step - loss: 0.1964 - accuracy: 0.9643 - val_loss: 0.0423 - val_accuracy: 0.9867
Epoch 221/300
22/21 [==============================] - 404s 18s/step - loss: 0.1996 - accuracy: 0.9415 - val_loss: 0.0417 - val_accuracy: 0.9867
Epoch 222/300
22/21 [==============================] - 395s 18s/step - loss: 0.1773 - accuracy: 0.9615 - val_loss: 0.0414 - val_accuracy: 0.9867
Epoch 223/300
22/21 [==============================] - 376s 17s/step - loss: 0.2102 - accuracy: 0.9601 - val_loss: 0.0417 - val_accuracy: 0.9867
Epoch 224/300
22/21 [==============================] - 383s 17s/step - loss: 0.2043 - accuracy: 0.9572 - val_loss: 0.0405 - val_accuracy: 0.9867
Epoch 225/300
22/21 [==============================] - 408s 19s/step - loss: 0.2149 - accuracy: 0.9601 - val_loss: 0.0397 - val_accuracy: 0.9867
Epoch 226/300
22/21 [==============================] - 406s 18s/step - loss: 0.1961 - accuracy: 0.9586 - val_loss: 0.0414 - val_accuracy: 0.9867
Epoch 227/300
22/21 [==============================] - 402s 18s/step - loss: 0.1673 - accuracy: 0.9629 - val_loss: 0.0415 - val_accuracy: 0.9867
Epoch 228/300
22/21 [==============================] - 402s 18s/step - loss: 0.1591 - accuracy: 0.9586 - val_loss: 0.0416 - val_accuracy: 0.9867
Epoch 229/300
22/21 [==============================] - 379s 17s/step - loss: 0.1636 - accuracy: 0.9715 - val_loss: 0.0414 - val_accuracy: 0.9867
Epoch 230/300
22/21 [==============================] - 377s 17s/step - loss: 0.1692 - accuracy: 0.9601 - val_loss: 0.0407 - val_accuracy: 0.9867
Epoch 231/300
22/21 [==============================] - 377s 17s/step - loss: 0.1970 - accuracy: 0.9586 - val_loss: 0.0408 - val_accuracy: 0.9867
Epoch 232/300
22/21 [==============================] - 389s 18s/step - loss: 0.2229 - accuracy: 0.9486 - val_loss: 0.0394 - val_accuracy: 0.9867
Epoch 233/300
22/21 [==============================] - 405s 18s/step - loss: 0.1558 - accuracy: 0.9615 - val_loss: 0.0394 - val_accuracy: 0.9867
Epoch 234/300
22/21 [==============================] - 405s 18s/step - loss: 0.1449 - accuracy: 0.9615 - val_loss: 0.0392 - val_accuracy: 0.9867
Epoch 235/300
22/21 [==============================] - 404s 18s/step - loss: 0.2053 - accuracy: 0.9558 - val_loss: 0.0389 - val_accuracy: 0.9867
Epoch 236/300
22/21 [==============================] - 402s 18s/step - loss: 0.1271 - accuracy: 0.9658 - val_loss: 0.0389 - val_accuracy: 0.9867
Epoch 237/300
22/21 [==============================] - 404s 18s/step - loss: 0.1817 - accuracy: 0.9572 - val_loss: 0.0397 - val_accuracy: 0.9867
Epoch 238/300
22/21 [==============================] - 404s 18s/step - loss: 0.1496 - accuracy: 0.9686 - val_loss: 0.0398 - val_accuracy: 0.9867
Epoch 239/300
22/21 [==============================] - 409s 19s/step - loss: 0.1831 - accuracy: 0.9672 - val_loss: 0.0387 - val_accuracy: 0.9867
Epoch 240/300
22/21 [==============================] - 414s 19s/step - loss: 0.1413 - accuracy: 0.9700 - val_loss: 0.0385 - val_accuracy: 0.9867
Epoch 241/300
22/21 [==============================] - 423s 19s/step - loss: 0.1639 - accuracy: 0.9643 - val_loss: 0.0383 - val_accuracy: 0.9867
Epoch 242/300
22/21 [==============================] - 409s 19s/step - loss: 0.1129 - accuracy: 0.9672 - val_loss: 0.0382 - val_accuracy: 0.9867
Epoch 243/300
22/21 [==============================] - 426s 19s/step - loss: 0.1635 - accuracy: 0.9700 - val_loss: 0.0385 - val_accuracy: 0.9867
Epoch 244/300
22/21 [==============================] - 410s 19s/step - loss: 0.1603 - accuracy: 0.9558 - val_loss: 0.0378 - val_accuracy: 0.9867
Epoch 245/300
22/21 [==============================] - 414s 19s/step - loss: 0.1672 - accuracy: 0.9544 - val_loss: 0.0379 - val_accuracy: 0.9867
Epoch 246/300
22/21 [==============================] - 407s 19s/step - loss: 0.1782 - accuracy: 0.9601 - val_loss: 0.0375 - val_accuracy: 0.9867
Epoch 247/300
22/21 [==============================] - 408s 19s/step - loss: 0.1594 - accuracy: 0.9601 - val_loss: 0.0367 - val_accuracy: 0.9867
Epoch 248/300
22/21 [==============================] - 407s 19s/step - loss: 0.1425 - accuracy: 0.9658 - val_loss: 0.0361 - val_accuracy: 0.9867
Epoch 249/300
22/21 [==============================] - 408s 19s/step - loss: 0.1300 - accuracy: 0.9658 - val_loss: 0.0363 - val_accuracy: 0.9867
Epoch 250/300
22/21 [==============================] - 407s 19s/step - loss: 0.1535 - accuracy: 0.9729 - val_loss: 0.0363 - val_accuracy: 0.9867
Epoch 251/300
22/21 [==============================] - 407s 19s/step - loss: 0.1406 - accuracy: 0.9629 - val_loss: 0.0352 - val_accuracy: 0.9867
Epoch 252/300
22/21 [==============================] - 408s 19s/step - loss: 0.1816 - accuracy: 0.9586 - val_loss: 0.0344 - val_accuracy: 0.9867
Epoch 253/300
22/21 [==============================] - 403s 18s/step - loss: 0.0873 - accuracy: 0.9729 - val_loss: 0.0339 - val_accuracy: 0.9867
Epoch 254/300
22/21 [==============================] - 405s 18s/step - loss: 0.1713 - accuracy: 0.9586 - val_loss: 0.0334 - val_accuracy: 0.9900
Epoch 255/300
22/21 [==============================] - 380s 17s/step - loss: 0.1089 - accuracy: 0.9786 - val_loss: 0.0327 - val_accuracy: 0.9900
Epoch 256/300
22/21 [==============================] - 388s 18s/step - loss: 0.1547 - accuracy: 0.9629 - val_loss: 0.0327 - val_accuracy: 0.9900
Epoch 257/300
22/21 [==============================] - 404s 18s/step - loss: 0.0888 - accuracy: 0.9658 - val_loss: 0.0331 - val_accuracy: 0.9900
Epoch 258/300
22/21 [==============================] - 405s 18s/step - loss: 0.1542 - accuracy: 0.9601 - val_loss: 0.0328 - val_accuracy: 0.9900
Epoch 259/300
22/21 [==============================] - 404s 18s/step - loss: 0.1545 - accuracy: 0.9643 - val_loss: 0.0321 - val_accuracy: 0.9900
Epoch 260/300
22/21 [==============================] - 404s 18s/step - loss: 0.1093 - accuracy: 0.9700 - val_loss: 0.0322 - val_accuracy: 0.9867
Epoch 261/300
22/21 [==============================] - 405s 18s/step - loss: 0.2043 - accuracy: 0.9572 - val_loss: 0.0323 - val_accuracy: 0.9867
Epoch 262/300
22/21 [==============================] - 406s 18s/step - loss: 0.0805 - accuracy: 0.9800 - val_loss: 0.0321 - val_accuracy: 0.9867
Epoch 263/300
22/21 [==============================] - 408s 19s/step - loss: 0.1556 - accuracy: 0.9686 - val_loss: 0.0321 - val_accuracy: 0.9900
Epoch 264/300
22/21 [==============================] - 404s 18s/step - loss: 0.1495 - accuracy: 0.9601 - val_loss: 0.0316 - val_accuracy: 0.9900
Epoch 265/300
22/21 [==============================] - 404s 18s/step - loss: 0.1125 - accuracy: 0.9700 - val_loss: 0.0318 - val_accuracy: 0.9867
Epoch 266/300
22/21 [==============================] - 404s 18s/step - loss: 0.1254 - accuracy: 0.9715 - val_loss: 0.0315 - val_accuracy: 0.9900
Epoch 267/300
22/21 [==============================] - 405s 18s/step - loss: 0.1122 - accuracy: 0.9672 - val_loss: 0.0303 - val_accuracy: 0.9900
Epoch 268/300
22/21 [==============================] - 404s 18s/step - loss: 0.1786 - accuracy: 0.9629 - val_loss: 0.0281 - val_accuracy: 0.9900
Epoch 269/300
22/21 [==============================] - 394s 18s/step - loss: 0.1488 - accuracy: 0.9629 - val_loss: 0.0284 - val_accuracy: 0.9900
Epoch 270/300
22/21 [==============================] - 376s 17s/step - loss: 0.1153 - accuracy: 0.9672 - val_loss: 0.0271 - val_accuracy: 0.9900
Epoch 271/300
22/21 [==============================] - 376s 17s/step - loss: 0.1501 - accuracy: 0.9672 - val_loss: 0.0263 - val_accuracy: 0.9900
Epoch 272/300
22/21 [==============================] - 376s 17s/step - loss: 0.1116 - accuracy: 0.9772 - val_loss: 0.0260 - val_accuracy: 0.9900
Epoch 273/300
22/21 [==============================] - 377s 17s/step - loss: 0.1853 - accuracy: 0.9658 - val_loss: 0.0261 - val_accuracy: 0.9900
Epoch 274/300
22/21 [==============================] - 403s 18s/step - loss: 0.1141 - accuracy: 0.9772 - val_loss: 0.0266 - val_accuracy: 0.9900
Epoch 275/300
22/21 [==============================] - 386s 18s/step - loss: 0.0824 - accuracy: 0.9772 - val_loss: 0.0267 - val_accuracy: 0.9900
Epoch 276/300
22/21 [==============================] - 378s 17s/step - loss: 0.1131 - accuracy: 0.9757 - val_loss: 0.0265 - val_accuracy: 0.9900
Epoch 277/300
22/21 [==============================] - 399s 18s/step - loss: 0.0716 - accuracy: 0.9772 - val_loss: 0.0265 - val_accuracy: 0.9900
Epoch 278/300
22/21 [==============================] - 406s 18s/step - loss: 0.0860 - accuracy: 0.9772 - val_loss: 0.0261 - val_accuracy: 0.9900
Epoch 279/300
22/21 [==============================] - 409s 19s/step - loss: 0.1323 - accuracy: 0.9658 - val_loss: 0.0263 - val_accuracy: 0.9900
Epoch 280/300
22/21 [==============================] - 415s 19s/step - loss: 0.1375 - accuracy: 0.9729 - val_loss: 0.0262 - val_accuracy: 0.9900
Epoch 281/300
22/21 [==============================] - 410s 19s/step - loss: 0.0969 - accuracy: 0.9757 - val_loss: 0.0258 - val_accuracy: 0.9900
Epoch 282/300
22/21 [==============================] - 408s 19s/step - loss: 0.1487 - accuracy: 0.9715 - val_loss: 0.0257 - val_accuracy: 0.9900
Epoch 283/300
22/21 [==============================] - 407s 18s/step - loss: 0.1220 - accuracy: 0.9729 - val_loss: 0.0256 - val_accuracy: 0.9900
Epoch 284/300
22/21 [==============================] - 396s 18s/step - loss: 0.1316 - accuracy: 0.9601 - val_loss: 0.0256 - val_accuracy: 0.9900
Epoch 285/300
22/21 [==============================] - 423s 19s/step - loss: 0.1575 - accuracy: 0.9601 - val_loss: 0.0265 - val_accuracy: 0.9900
Epoch 286/300
22/21 [==============================] - 420s 19s/step - loss: 0.1653 - accuracy: 0.9672 - val_loss: 0.0261 - val_accuracy: 0.9900
Epoch 287/300
22/21 [==============================] - 409s 19s/step - loss: 0.1240 - accuracy: 0.9686 - val_loss: 0.0253 - val_accuracy: 0.9900
Epoch 288/300
22/21 [==============================] - 410s 19s/step - loss: 0.1148 - accuracy: 0.9700 - val_loss: 0.0251 - val_accuracy: 0.9900
Epoch 289/300
22/21 [==============================] - 408s 19s/step - loss: 0.1235 - accuracy: 0.9800 - val_loss: 0.0255 - val_accuracy: 0.9900
Epoch 290/300
22/21 [==============================] - 407s 18s/step - loss: 0.1931 - accuracy: 0.9643 - val_loss: 0.0243 - val_accuracy: 0.9900
Epoch 291/300
22/21 [==============================] - 408s 19s/step - loss: 0.1951 - accuracy: 0.9700 - val_loss: 0.0244 - val_accuracy: 0.9900
Epoch 292/300
22/21 [==============================] - 386s 18s/step - loss: 0.0943 - accuracy: 0.9786 - val_loss: 0.0240 - val_accuracy: 0.9900
Epoch 293/300
22/21 [==============================] - 383s 17s/step - loss: 0.2295 - accuracy: 0.9658 - val_loss: 0.0238 - val_accuracy: 0.9900
Epoch 294/300
22/21 [==============================] - 380s 17s/step - loss: 0.1028 - accuracy: 0.9743 - val_loss: 0.0238 - val_accuracy: 0.9900
Epoch 295/300
22/21 [==============================] - 387s 18s/step - loss: 0.1162 - accuracy: 0.9672 - val_loss: 0.0236 - val_accuracy: 0.9900
Epoch 296/300
22/21 [==============================] - 381s 17s/step - loss: 0.0669 - accuracy: 0.9800 - val_loss: 0.0233 - val_accuracy: 0.9900
Epoch 297/300
22/21 [==============================] - 379s 17s/step - loss: 0.1472 - accuracy: 0.9658 - val_loss: 0.0230 - val_accuracy: 0.9900
Epoch 298/300
22/21 [==============================] - 379s 17s/step - loss: 0.1105 - accuracy: 0.9757 - val_loss: 0.0222 - val_accuracy: 0.9900
Epoch 299/300
22/21 [==============================] - 380s 17s/step - loss: 0.1339 - accuracy: 0.9686 - val_loss: 0.0212 - val_accuracy: 0.9900
Epoch 300/300
22/21 [==============================] - 379s 17s/step - loss: 0.0940 - accuracy: 0.9786 - val_loss: 0.0197 - val_accuracy: 0.9900

标签： pythonimagekerasdeep-learningcomputer-vision