ValueError：检查目标时出错：预期 activation_17 具有 2 维，但得到的数组形状为 (1, 256, 256, 3)

Question

我正在构建一个基于 VGG-16 预测人类年龄的 CNN，并输入两个 RGB 图像。 （文件类型：jpg）。我在 Python 2.7 的 anaconda 环境中使用 tensorflow 后端。

但是，它总是会引发错误：

Traceback（最近一次调用最后一次）：

文件“train2.py”，第 167 行，在 洗牌=真）

文件“/Users/name/anaconda/lib/python2.7/sitepackages/keras/models.py”，第 973 行，适合 validation_steps=validation_steps）

文件“/Users/name/anaconda/lib/python2.7/sitepackages/keras/engine/training.py”，第 1581 行，适合 batch_size=batch_size)

文件“/Users/name/anaconda/lib/python2.7/sitepackages/keras/engine/training.py”，第 1418 行，在 _standardize_user_data exception_prefix='target')

文件“/Users/name/anaconda/lib/python2.7/sitepackages/keras/engine/training.py”，第 141 行，在 _standardize_input_data str(array.shape))

ValueError：检查目标时出错：预期activation_17 到

有 2 个维度，但得到了形状为 (1, 256, 256, 3) 的数组

我该如何解决这个错误？这是代码：

import keras

from keras.preprocessing.image import ImageDataGenerator

from keras.models import Sequential

from keras.layers import Dense

from keras.layers import Dropout

from keras.layers import Activation

from keras.layers import Flatten

from keras.layers import Conv2D

from keras.layers import MaxPooling2D

from keras.layers import BatchNormalization

from keras.utils import np_utils

import cv2

from PIL import Image

import numpy as np

from sklearn.cross_validation import train_test_split

from sklearn.model_selection import train_test_split

name_path = ["pathname"]

new_age_list=[45,52]

img_rows=256

img_cols=256

img_array = 
np.array([np.array(Image.open(i).resize((img_rows,img_cols),Image.BILINEAR)) for i in name_path[0:2]],"f")

(X, y) = (img_array[0:2],new_age_list[0:2])

y=np.asarray(y)

X=X.reshape(2,256,256,3)

# STEP 1: split X and y into training and testing sets

train_data, train_label,test_data, test_label= train_test_split(X, y, 
test_size=0.5, random_state=4)

train_data = train_data.astype('float32')

test_data = test_data.astype('float32')

train_data = train_data / 255

test_data = test_data / 255

""" Model """

model = Sequential()

""" Block 1 """

model.add(Conv2D(64, (3,3), padding='same', 
border_mode='valid',input_shape=(256,256,3)))

model.add(BatchNormalization())

model.add(Activation('relu'))

model.add(Conv2D(64, (3,3)))

model.add(BatchNormalization())

model.add(Activation('relu'))

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

""" Block 2 """

model.add(Conv2D(128, (3,3)))

model.add(BatchNormalization())

model.add(Activation('relu'))

model.add(Conv2D(128, (3,3)))

model.add(BatchNormalization())

model.add(Activation('relu'))

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

""" Block 3 """

model.add(Conv2D(256, (3,3)))

model.add(BatchNormalization())

model.add(Activation('relu'))

model.add(Conv2D(256, (3,3)))

model.add(BatchNormalization())

model.add(Activation('relu'))

model.add(Conv2D(256, (3,3)))

model.add(BatchNormalization())

model.add(Activation('relu'))

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

""" Block 4 """
model.add(Conv2D(512, (3,3)))

model.add(BatchNormalization())

model.add(Activation('relu'))

model.add(Conv2D(512, (3,3)))

model.add(BatchNormalization())

model.add(Activation('relu'))

model.add(Conv2D(512, (3,3)))

model.add(BatchNormalization())

model.add(Activation('relu'))

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

""" Block 5 """

model.add(Conv2D(512, (3,3)))

model.add(BatchNormalization())

model.add(Activation('relu'))

model.add(Conv2D(512, (3,3)))

model.add(BatchNormalization())

model.add(Activation('relu'))

model.add(Conv2D(512, (3,3)))

model.add(BatchNormalization())

model.add(Activation('relu'))

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

model.add(Dropout(0.25))

""" Flatten """

model.add(Flatten())

model.add(Dense(512))

model.add(BatchNormalization())

model.add(Activation('relu'))

model.add(Dense(128))

model.add(BatchNormalization())
model.add(Activation('relu'))

model.add(Dense(32))

model.add(BatchNormalization())

model.add(Activation('relu'))

model.add(Dense(1))

model.add(BatchNormalization())

model.add(Activation('softmax'))

""" Optimizer """

opt = keras.optimizers.rmsprop(lr=config.learning_rate, 

decay=config.decay)

print model.summary()

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

""" Fit Data """

batch_size = 512

epoch = 1000

learning_rate = 1e-4

decay = 1e-7

for i in range(epoch):

    model.fit(train_data, train_label,

              batch_size=batch_size,

              epochs=int(epoch/epoch),

              validation_data=(test_data, test_label),

              shuffle=True)

Answer 1

我假设“activation_17”是指最后一个激活层，即“softmax”层。由于您的图像可以毫无问题地通过图层传播，我认为标签的尺寸存在问题。

在检查了documentation 中的sklearn.model_selection.train_test_split 之后，您似乎在分配train_test_split() 的返回值时犯了一个错误。

像这样尝试：

train_data, test_data, train_label, test_label= train_test_split(X, y, test_size=0.5, random_state=4)