RapidLib

RapidLib is a lightweight library for interactive machine learning, inspired by Wekinator. It currently features classification (using kNN), regression (multilayer perceptron), and series classification (using dynamic time warping). It also includes rapidStream, a collection of utilities for buffering and signal processing on incoming streams of data.

C++

Library

Full C++ documentation can be found here.

OpenFrameworks addon

https://github.com/mzed/ofxRapidLib

Javascript

Embeded Library

RapidLib has also been transpiled to JavaScript using Emscripten. It can be used in a browser through a script tag.

<script src="https://www.doc.gold.ac.uk/eavi/rapidmix/RapidLib.js"></script>

Node module

npm install rapidlib

Full JavaScript documentation can be found here

Max objects

RapidLib has been wrapped in a Max object here.

Game Engines

RapidLib is part of InteractML, for Unity and Unreal Engine

MIMIC

RapidLib is part of the code for Learner.js in the MIMIC project.

Here is a guide to using RapidLib on the MIMIC platform.