How can I visualize the weights(variables) in cnn in Tensorflow?

by

To visualize the weights, you can use a tf.image_summary() op to transform a convolutional filter (or a slice of a filter) into a summary proto, write them to a log using a tf.train.SummaryWriter, and visualize the log using TensorBoard.

Let’s say you have the following (simplified) program:

filter = tf.Variable(tf.truncated_normal([8, 8, 3]))
images = tf.placeholder(tf.float32, shape=[None, 28, 28])

conv = tf.nn.conv2d(images, filter, strides=[1, 1, 1, 1], padding="SAME")

# More ops...
loss = ...
optimizer = tf.GradientDescentOptimizer(0.01)
train_op = optimizer.minimize(loss)

filter_summary = tf.image_summary(filter)

sess = tf.Session()
summary_writer = tf.train.SummaryWriter('/tmp/logs', sess.graph_def)
for i in range(10000):
  sess.run(train_op)
  if i % 10 == 0:
    # Log a summary every 10 steps.
    summary_writer.add_summary(filter_summary, i)

After doing this, you can start TensorBoard to visualize the logs in /tmp/logs, and you will be able to see a visualization of the filter.

Note that this trick visualizes depth-3 filters as RGB images (to match the channels of the input image). If you have deeper filters, or they don’t make sense to interpret as color channels, you can use the tf.split() op to split the filter on the depth dimension, and generate one image summary per depth.

More Related Contents:

  1. What are logits? What is the difference between softmax and softmax_cross_entropy_with_logits?
  2. TensorFlow Only running on 1/32 of the Training data provided [duplicate]
  3. How to get current available GPUs in tensorflow?
  4. How to export Keras .h5 to tensorflow .pb?
  5. How to choose cross-entropy loss in TensorFlow?
  6. Keras + Tensorflow and Multiprocessing in Python
  7. Is it still necessary to install CUDA before using the conda tensorflow-gpu package?
  8. Meaning of inter_op_parallelism_threads and intra_op_parallelism_threads
  9. How do I use TensorFlow GPU?
  10. Is Tensorflow compatible with a Windows workflow?
  11. Tensorflow, best way to save state in RNNs?
  12. Remove nodes from graph or reset entire default graph
  13. How does Keras 1d convolution layer work with word embeddings – text classification problem? (Filters, kernel size, and all hyperparameter)
  14. Streaming large training and test files into Tensorflow’s DNNClassifier
  15. What do I need K.clear_session() and del model for (Keras with Tensorflow-gpu)?
  16. Clearing Tensorflow GPU memory after model execution
  17. Tensorflow: Using Adam optimizer
  18. Why can’t I get reproducible results in Keras even though I set the random seeds?
  19. How to get stable results with TensorFlow, setting random seed
  20. Tensor is not an element of this graph
  21. overcome Graphdef cannot be larger than 2GB in tensorflow
  22. Training a simple model in Tensorflow GPU slower than CPU
  23. How to fix “AttributeError: module ‘tensorflow’ has no attribute ‘get_default_graph'”?
  24. What is right batch normalization function in Tensorflow?
  25. Why is numpy native on M1 Max greatly slower than on old Intel i5?
  26. Is it thread-safe when using tf.Session in inference service?
  27. Strange behaviour of the loss function in keras model, with pretrained convolutional base
  28. Attach a queue to a numpy array in tensorflow for data fetch instead of files?
  29. TensorFlow 2.0 Keras: How to write image summaries for TensorBoard
  30. Multi-input Multi-output Model with Keras Functional API

Leave a Comment