Software Platform

Seniorforsker
920 39 217

The most time consuming, and typically least enjoyable tasks in data science are cleaning, synchronising and organizing data (Source: CrowdFlowerDataScientist report 2017). This is caused by a tool-gap as there are few established tools available in a research setting. We have therefore developed the AutoActive Research Environment as a part of the AutoActive project which aims to make data science in a research environment easier.

The AutoActive Research Environment includes ActivityPresenter which is an easy-to-use graphical user interface and two toolboxes for Matlab and Python. ActivityPresenter supports:

  • Importing data from a variety of sources such as:
      • GaitUp and Catapult sensors
      • Common file formats such as comma separated values (.csv) and excel (.xlsx)
      • Video files
  • Visualising sensor data and videos
  • Synchronising data from multiple sources such as
      • Videos and sensors
      • Different sensor sources
      • Different video sources
  • Organizing data from multiple sources into a single file (.aaz) which is easy to read with the developed Matlab and Python toolboxes.

The goal of the Matlab and Python toolboxes is to make it possible to read and write aaz files. This makes it possible to use all the powerful algorithms already implemented in Python and Matlab, but at the same time use the graphical user interface to visualise, synchronise and organise the data.

ActivityPresenter is available here as a Microsoft store app

Furthermore, the source code for ActivityPresenter, the Matlab toolbox and the Python toolbox is available as Open Source in these repositories:

Documentation and examples are also available in the respective GitHub repositories.