如何转换“张量"到“numpy"张量流中的数组? [英] How to convert "tensor" to "numpy" array in tensorflow?

问题描述

我试图在 tesnorflow2.0 版本中将张量转换为 numpy.由于 tf2.0 启用了急切执行，因此它应该默认工作并且在正常运行时也工作.当我在 tf.data.Dataset API 中执行代码时，它给出了一个错误

I am trying to convert a tensor to numpy in the tesnorflow2.0 version. Since tf2.0 have eager execution enabled then it should work by default and working too in normal runtime. While I execute code in tf.data.Dataset API then it gives an error

"AttributeError: 'Tensor' 对象没有属性 'numpy'"

"AttributeError: 'Tensor' object has no attribute 'numpy'"

我在 tensorflow 变量之后尝试了.numpy()"，而对于.eval()"，我无法获得默认会话.

I have tried ".numpy()" after tensorflow variable and for ".eval()" I am unable to get default session.

from __future__ import absolute_import, division, print_function, unicode_literals
import tensorflow as tf
# tf.executing_eagerly()
import os
import time
import matplotlib.pyplot as plt
from IPython.display import clear_output
from model.utils import  get_noise
import cv2


def random_noise(input_image):
  img_out = get_noise(input_image)
  return img_out


def load_denoising(image_file):
  image = tf.io.read_file(image_file)
  image = tf.image.decode_png(image)
  real_image = image
  input_image = random_noise(image.numpy())
  input_image = tf.cast(input_image, tf.float32)
  real_image = tf.cast(real_image, tf.float32)
  return input_image, real_image


def load_image_train(image_file):
  input_image, real_image = load_denoising(image_file)
  return input_image, real_image

这很好用

inp, re = load_denoising('/data/images/train/18.png')
# Check for correct run
plt.figure()
plt.imshow(inp)
print(re.shape,"  ", inp.shape)

这会产生提到的错误

train_dataset = tf.data.Dataset.list_files('/data/images/train/*.png')
train_dataset = train_dataset.map(load_image_train,num_parallel_calls=tf.data.experimental.AUTOTUNE)

注意:random_noise 有 cv2 和 sklearn 函数

Note: random_noise have cv2 and sklearn functions

推荐答案

你不能在张量上使用 .numpy 方法，如果这个张量要在 中使用tf.data.Dataset.map 调用.

You can't use the .numpy method on a tensor, if this tensor is going to be used in a tf.data.Dataset.map call.

引擎盖下的 tf.data.Dataset 对象通过创建静态图来工作:这意味着您不能使用 .numpy() 因为 tf.Tensor 对象在静态图上下文中没有此属性.

The tf.data.Dataset object under the hood works by creating a static graph: this means that you can't use .numpy() because the tf.Tensor object when in a static-graph context do not have this attribute.

因此，行 input_image = random_noise(image.numpy()) 应该是 input_image = random_noise(image).

Therefore, the line input_image = random_noise(image.numpy()) should be input_image = random_noise(image).

但是由于random_noise 调用了model.utils 包中的get_noise，代码很可能再次失败.如果 get_noise 函数是使用 Tensorflow 编写的，那么一切都会正常进行.否则，它将无法工作.

But the code is likely to fail again since random_noise calls get_noise from the model.utils package. If the get_noise function is written using Tensorflow, then everything will work. Otherwise, it won't work.

解决方案?仅使用 Tensorflow 原语编写代码.

The solution? Write the code using only the Tensorflow primitives.

例如，如果您的函数 get_noise 只是使用输入图像的薄片创建随机噪声，您可以将其定义为:

For instance, if your function get_noise just creates random noise with the shee of your input image, you can define it like:

def get_noise(image):
    return tf.random.normal(shape=tf.shape(image))

只使用 Tensorflow 原语，它会起作用.

using only the Tensorflow primitives, and it will work.

希望这篇概述能有所帮助！

Hope this overview helps!

PS:您可能有兴趣查看文章分析 tf.function 以发现 AutoGraph 的优势和微妙之处"-它们涵盖了这方面(也许第 3 部分与您的场景相关):第 1 部分 第 2 部分 第 3 部分

P.S: you could be interested in having a look at the articles "Analyzing tf.function to discover AutoGraph strengths and subtleties" - they cover this aspect (perhaps part 3 is the one related to your scenario): part 1 part 2 part 3

这篇关于如何转换“张量"到“numpy"张量流中的数组?的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！