Open Source Deep Learning in Java – Deep Netts Community Edition

What is Deep Netts Community Edition?

• A free open-source version of Deep Netts Java deep learning library
• Pure Java implementation of the most important and commonly used deep learning algorithms
• Used as a reference implementation for the official Java technology standard for Visual Recognition API, JSR381
• Licensed under open-source GNU GPLv2 with the Classpath Exception. For more details see here
• It is suitable to try tutorial examples and learn basics of deep learning in Java.

How to Get and Use Deep Netts Community Edition?

• Deep Netts Community Edition is available on Central Maven Repository. Dependency for Deep Netts to use for Maven-based Java project:

<dependency>
     <groupId>com.deepnetts</groupId>
     <artifactId>deepnetts-core</artifactId>
     <version>1.13.2</version>
</dependency>

• Code examples using Deep Netts Community Edition are available on the GitHub: https://github.com/deepnetts/deepnetts-communityedition/tree/community-visrec/deepnetts-examples
• Source code of the Deep Netts Community edition is also available on the GitHub: https://github.com/deepnetts/deepnetts-communityedition

What are the differences between Community and Professional Editions?

Professional Edition provides

• Tools: Visual development tools with project templates, best-practice workflows and expert assistant
• Better Accuracy and advanced features more advanced algorithms and additional algorithm features
• Increased performance. Over 3X faster depending of the size of network and CPU.
• Premium support: documentation, updates, fixes

Professional Edition can be used free for development and low volume production. Free Download is available here