This is machine translation

Translated by Microsoft
Mouseover text to see original. Click the button below to return to the English version of the page.

Note: This page has been translated by MathWorks. Click here to see
To view all translated materials including this page, select Country from the country navigator on the bottom of this page.

Deep Network Designer

Edit and build deep learning networks

Description

The Deep Network Designer app lets you build, visualize, and edit deep learning networks. Using this app, you can:

  • Import pretrained networks and edit them for transfer learning.

  • Import and edit networks, and build new networks.

  • Drag and drop to add new layers and create new connections.

  • View and edit layer properties.

  • Analyze the network to ensure you define the architecture correctly, and detect problems before training.

After you finish designing a network, you can export it to the workspace where you can save or train the network.

Open the Deep Network Designer App

  • MATLAB® Toolstrip: On the Apps tab, under Machine Learning and Deep Learning, click the app icon.

  • MATLAB command prompt: Enter deepNetworkDesigner.

Examples

expand all

Examine a simple pretrained network in the app.

Load a simple pretrained network. If you need to download the network, use the download link.

net = squeezenet

Open the app.

deepNetworkDesigner

In the File section, click Import and choose the network to load from the workspace.

Use the plot to explore and visualize the network.

For a list of available networks and how to compare them, see Pretrained Convolutional Neural Networks.

Edit a network in the app to prepare it for transfer learning.

Load a pretrained network. If you need to download the network, use the download link.

net = googlenet

Open the app.

deepNetworkDesigner

In the File section, click Import and choose the network to load from the workspace. Use the plot to explore and visualize the network.

Edit the network to specify a new number of classes in your data. Drag a new fully connected layer onto the canvas and edit OutputSize property to a new number of classes. Delete the last fully connected layer and connect up your new layer instead.

Delete the classification output layer. Then, drag a new classification output layer onto the canvas and connect it instead. The output layer auto settings will learn the number of classes during training.

To check that the network is ready for training, click Analyze in the Analysis section.

Return to the Deep Network Designer. To export the network to the workspace for training, in the Export section, click Export.

For more help, see Transfer Learning with Deep Network Designer.

For help understanding and editing layer properties, consult the layer pages.

To find definitions and help on all layer properties, click the layer name in the table List of Deep Learning Layers.

In the app, click layers to view and edit properties.

Related Examples

See Also

Functions

Introduced in R2018b