One-click Installation: Unzip the ZIP file and drag-and-drop the 5 JAR files into ImageJ or Fiji. Restart Fiji.
How to use: Check the Documentation of the plugin.
Source code: GitHub Java source code of the plugin.
Older releases: The current version of the plugin is built with TensorFlow Java API version 1.15, but it is possible to download releases that use version 1.12.
DeepImageJ is split into three different modules:
DeepImageJ Run: Applies a convolutional neural network to an input image (it is macro-recordable).
DeepImageJ Explore: Allows to explore all the installed models.
DeepImageJ BuildBundledModel: A guided user interface (GUI) to build bundled models from TensorFlow models.
Saving Keras models to be loaded in DeepImageJ: Python code and examples in GitHub.
The DeepImageJ project is an open-source software (OSS) under the BSD 2-Clause License. All the resources provided here are freely available. As a matter of academic integrity, we strongly encourage users to include adequate references whenever they present or publish results that are based on the resources provided here.