The repository contains the source code of the reference implementation of JSR 381, a standardization in Java for Visual Recognition.
The Visual Recognition API JSR #381 is a software development standard recognized by the Java Community Process (JCP) that simplifies and standardizes a set of APIs familiar to Java developers for classifying and recognizing objects in images using machine learning. Beside classes specific for visual recognition tasks, it provides general abstractions for machine learning tasks like classification, regression and data set, and reusable design which can be applied for machine learning systems in other domains. At the current stage it provides basic hello world examples for supported machine learning tasks (classification and regression) and image classification.
This reference implementation is based on Deep Java Library (DJL) available at
https://github.com/awslabs/djl
Specification for VisRec API is available at
https://github.com/JavaVisRec/visrec-api
For step by step guide and additional info see getting started guide at
https://github.com/JavaVisRec/visrec-api/wiki/Getting-Started-Guide