Skip to content

mano-js is a library targeted at sensor processing and gesture modeling and recognition

License

Notifications You must be signed in to change notification settings

Ircam-RnD/mano-js

Repository files navigation

mano-js

mano-js is a library targeted at sensor processing and gesture modeling and recognition. The library is designed to offer a high-level client-side wrapper of waves-lfo, lfo-motion, xmm-client.

Overview

scheme

Install

npm install [--save --save-exact] ircam-rnd/mano-js

Example

import * as mano from 'mano-js/client';

const processedSensors = new mano.ProcessedSensors();
const example = new mano.Example();
const trainingSet = new mano.TrainingSet();
const xmmProcessor = new mano.XmmProcesssor();

example.setLabel('test');
processedSensors.addListener(example.addElement);

// later...
processedSensors.removeListener(example.addElement);
const rapidMixJsonExample = example.toJSON();

trainingSet.addExample(rapidMixJsonExample);
const rapidMixJsonTrainingSet = trainingSet.toJSON();

xmmProcessor
  .train(rapidMixJsonTrainingSet)
  .then(() => {
    // start decoding
    processedSensors.addListener(data => {
      const results = xmmProcessor.run(data);
      console.log(results);
    });
  });

Server-side considerations

By default, the training is achieved by calling the dedicated service available at https://como.ircam.fr/api/v1/train, however a similar service can be simply deployed by using the xmm-node and rapid-mix adapters librairies.

An concrete example of such solution is available in examples/mano-js-example.

Acknowledgements

The library as been developped at Ircam - Centre Pompidou by Joseph Larralde and Benjamin Matuszewski in the framework of the EU H2020 project Rapid-Mix.

About

mano-js is a library targeted at sensor processing and gesture modeling and recognition

Resources

License

Stars

Watchers

Forks

Packages

No packages published