如何在tensorflow上可视化学习过滤器 [英] How to visualize learned filters on tensorflow

问题描述

类似于Caffe框架，在CNNs训练期间可以观察学习过滤器，并且由于输入图像而产生卷积，我想知道是否可以用TensorFlow做同样的事情？

一个Caffe示例可以在这个链接中查看：

http://nbviewer.jupyter.org/github/BVLC/caffe/blob/master/examples/ 00-classification.ipynb

感谢您的帮助！

解决方案

要在Tensorboard中查看几个conv1过滤器，您可以使用此代码（它适用于cifar10）

pre > ＃这应该是cifar10.py文​​件中的推理（图像）函数的一部分
$ b＃conv1
与tf.variable_scope（'conv1'）作为范围：
kernel = _variable_with_weight_decay（'weights'，shape = [5，5，3，64]，
stddev = 1e-4，wd = 0.0）
conv = tf.nn.conv2d（images，kernel，[1,1,1,1]，padding ='SAME'）
biasses = _variable_on_cpu 'bias'，[64]，tf.constant_initializer（0.0））
bias = tf.nn.bias_add（conv，biasses）
conv1 = tf.nn.relu（bias，name = scope.name ）
_activation_summary（conv1）

with tf.variable_scope（'visualization'）：
＃将权重缩放为[0 1]，类型仍为浮点数
x_min = tf.reduce_min（kernel）
x_max = tf.reduce_max（kernel）
kernel_0_to_1 =（kernel-x_min）/（x_max - x_min）

＃to tf.image_summary format [ batch_size，height，width，channels]
kernel_transposed = tf.transpose（kernel_0_to_1，[3,0,1,2]）

＃这将在conv1中显示随机的3个过滤器
tf.image_summary（'conv1 / filters'，kernel_transposed，max_images = 3）


I也写了一个简单的 gist 在网格中显示所有64个conv1过滤器。

Similarly to the Caffe framework, where it is possible to watch the learned filters during CNNs training and it's resulting convolution with input images, I wonder if is it possible to do the same with TensorFlow?

A Caffe example can be viewed in this link:

http://nbviewer.jupyter.org/github/BVLC/caffe/blob/master/examples/00-classification.ipynb

Grateful for your help!

解决方案

To see just a few conv1 filters in Tensorboard, you can use this code (it works for cifar10)

# this should be a part of the inference(images) function in cifar10.py file

# conv1
with tf.variable_scope('conv1') as scope:
  kernel = _variable_with_weight_decay('weights', shape=[5, 5, 3, 64],
                                       stddev=1e-4, wd=0.0)
  conv = tf.nn.conv2d(images, kernel, [1, 1, 1, 1], padding='SAME')
  biases = _variable_on_cpu('biases', [64], tf.constant_initializer(0.0))
  bias = tf.nn.bias_add(conv, biases)
  conv1 = tf.nn.relu(bias, name=scope.name)
  _activation_summary(conv1)

  with tf.variable_scope('visualization'):
    # scale weights to [0 1], type is still float
    x_min = tf.reduce_min(kernel)
    x_max = tf.reduce_max(kernel)
    kernel_0_to_1 = (kernel - x_min) / (x_max - x_min)

    # to tf.image_summary format [batch_size, height, width, channels]
    kernel_transposed = tf.transpose (kernel_0_to_1, [3, 0, 1, 2])

    # this will display random 3 filters from the 64 in conv1
    tf.image_summary('conv1/filters', kernel_transposed, max_images=3)

I also wrote a simple gist to display all 64 conv1 filters in a grid.

这篇关于如何在tensorflow上可视化学习过滤器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！