RAPID-MIX API already works with some useful and exciting tools that form part of the growing RAPID-MIX ecosystem:


RepoVizz2 is an online repository for multimodal datasets. It is useful for collaborative data-driven projects and supports structural formatting, remote storage, browsing, exchange and annotation of synchronous multimodal and time-aligned data.

Users can store, edit, and retrieve multimodal datasets online via a RESTful API.

Documentation & python tutorial available at https://github.com/repovizz/repovizz2doc

The RAPID-MIX API includes classes that allow users to share data via JSON to and from Machine Learning and Training Data classes.

Freesound API

Freesound is an online Collaborative database of Creative Commons licensed sounds.

  • Browse, search, and retrieve information about Freesound sounds, users, and packs
  • Find similar sounds to a given target (based on content analysis)
  • Retrieve automatically extracted features from audio files
  • Perform advanced queries combining content analysis features and other metadata (tags, etc…).

Latest release here: https://www.freesound.org/docs/api/overview.html


PiPo is a plugin interface and collection of processing objects that implement filtering, transformation, descriptor extraction, and segmentation algorithms that can be applied to audio and multi-dimensional streams of gesture and motion data, in real-time or offline. The SDK allows anybody to build their own PiPo operators, and offers base functionality for implementing hosts. Latest release here: http://ismm.ircam.fr/pipo/

Pipo is the underlying architecture for the RAPID-MIX API. The API can host PiPo plugins.


Wekinator for RAPID-MIX

The Wekinator allows users to build new interactive systems by demonstrating human actions and computer responses, instead of writing programming code. This version saves trained models to be used in embedded code, or imported into the RAPID-MIX API. It outputs both:

  • a C++ code version of current trained models.
  • a JSON description file that can be loaded into a RapidLib modelSet.

Latest release here: https://github.com/mzed/wekimini/releases

RapidLib for Unity

Marco Gilles has contributed a set of Unity bindings for RapidLib here: http://gitlab.doc.gold.ac.uk/mgill009/rapidlib-unity

RapidLib for Node

Here is a rapidlib node module. It can be installed with npm, like:

RapidLib for openFrameworks

An openFrameworks addon of RapidLib can be found here.