.. _example_plot_overfeat_layer1_filters.py:


====================================
Visualization of first layer filters
====================================

The first layers of convolutional neural networks often have very "human
interpretable" values, as seen in these example plots. Visually, these filters
are similar to other filters used in computer vision, such as Gabor filters.




.. rst-class:: horizontal


    *

      .. image:: images/plot_overfeat_layer1_filters_001.png
            :scale: 47

    *

      .. image:: images/plot_overfeat_layer1_filters_002.png
            :scale: 47




**Python source code:** :download:`plot_overfeat_layer1_filters.py <plot_overfeat_layer1_filters.py>`

.. literalinclude:: plot_overfeat_layer1_filters.py
    :lines: 11-

**Total running time of the example:**  0.40 seconds
( 0 minutes  0.40 seconds)