我正在尝试使用Reshape图层重构张量:
from keras.layers.convolutional import Conv2D, MaxPooling2D,AveragePooling2D
from keras import backend as K
from keras.models import Model
from keras.layers import Input
from keras.layers.core import Activation, Reshape
from keras.layers import Dense,Reshape,Lambda,Dropout
import numpy as np
from keras.layers.embeddings import Embedding
Dict_size=32
EmbedSz=16
img_sz=100
channels=3
input=Input(shape=(img_sz,img_sz,channels))
H=Conv2D(Dict_size, 3, 3, activation='relu', border_mode='same')(input)
H=Lambda(lambda x:K.argmax(x, axis=3),output_shape=lambda s: (img_sz,img_sz,))(H)
H=Reshape((1,img_sz*img_sz))(H)
model=Model(inputs=input,outputs=H)
#model.compile( optimizer= 'adam', metrics=[ 'accuracy' ],loss='mse')
ar=np.random.rand(1,100,100,3)
pr=model.predict(ar)
print(pr.shape)
print(pr)$
但得到这个错误!文件“/usr/local/lib/python2.7/dist-packages/keras/layers/core.py”,第379行,在_fix_unknown_dimension中引发ValueError(msg)ValueError:新数组的总大小必须保持不变
我没有改变大小!
1 回答
您只是忘了将批量大小维度添加到Lambda图层:
所以,只需将
None
添加到output_shape即可 .