tf.keras.layers.UpSampling2D()（tensorflow2.7.0）与tf.tile(tensorflow1.15.0)的区别

UpSampling2D

>>>input = numpy.array([[1,2,3],[4,5,6]])
>>>input_shape = Input(shape=(2,3,1))
>>>print(input)
[[1 2 3]
[4 5 6]]
>>>out = tf.keras.layers.UpSampling2D(size=(2,2))(input_shape)
>>>print(out)
[[1 1 2 2 3 3]
 [1 1 2 2 3 3]
 [4 4 5 5 6 6]
 [4 4 5 5 6 6]]

UpSampling2D将原矩阵分别沿着原来的数值阵列对应的倍数复制。

tf.tile()

tf.tile(
    input,#tensor输入
    multiples,#tensor维度大小的输入
    name=None
)

tf.tile应用

with tf.Session() as sess:
    a = tf.constant([[1,2,3],[4,5,6]])
    b = tf.tile(a,[2,2])
    print(sess.run(a))
    print(sess.run(b))

#output:
[[1 2 3]
 [4 5 6]]
[[1 2 3 1 2 3]
 [4 5 6 4 5 6]
 [1 2 3 1 2 3]
 [4 5 6 4 5 6]]
 

tile有平铺的意思，进行同一维度上的复制。