VCBot

The Virtual Cartesian Robot application for human-robot interaction.

Introduction

The virtual Cartesian robot (VCBot) is a program designed for interacting with an artificial neural agent through the computer mouse. VCBot is implemented in the Python programming language running on the top of the neural robotics library (see the NRL project), which is an open-source tool implemented in the C++ programming language for prototyping human-robot interaction experiments, based on artificial neural cognitive control from recurrent neural network models.

Project content

In this repository the implementation of VCBot is provided in the Python programming language version 3.7. The graphical user interface (GUI) relies on the tkinter toolkit, so the program can run in Linux, Microsoft Windows, and Mac OS X platforms. A user guide is provided here with details on how to operate the GUI. The tutorial video below presents the project and explains how to run VCBot .

The instructions to install Python and the tkinter toolkit in multiple platforms are provided next.

Requirements

Please download the NRL project, follow the instruction provided for compiling the library in your host platform, and place the library file in the folder src/lib.

VCBot was developed in Python version 3.7, It requires the following packages: numpy, sklearn, pandas, parse, matplotlib, IPython, and ttkthemes

Installation

It is recommended to create a virtual environment for running the client application, in order to preserve previous configurations in your system.

Ubuntu 16.04

• Install the tkinter toolkit

  sudo apt-get install python3.7-tk
  

• Create a virtual environment (recommended) Install the virtual environment

  sudo pip3 install virtualenv
  virtualenv --python=/usr/bin/python3.7 ~/Workspace/virtual3_7
  

  Activate the virtual environment

  source DESIRED_PATH/virtual3_7/bin/activate
  

Windows 8.1-10.x

• Install the tkinter toolkit

Download Python from https://www.python.org/downloads/windows/ The version Python 3.7.3 - March 25, 2019 was successfully tested. In advanced options select install tkinter and add python to the PATH environment variable .

• Create a virtual environment (recommended)

  Install the virtual environment

  pip install virtualenv
  virtualenv DESIRED_PATH\virtual3_7
  

  Activate the virtual environment

  DESIRED_PATH/virtual3_7/Scripts/activate.bat
  

MAC OS X

• install homebrew

• Install the tkinter toolkit

  brew install tcl-tk
  

• Create a virtual environment (optional but recommended) Install the virtual environment

  virtualenv --python=/usr/bin/python3.7 DESIRED_PATH/virtual3_7
  

  Activate the virtual environment

  source ~/Workspace/virtual3_7/bin/activate
  

All platforms

• Install the Python required packages

  pip install --upgrade setuptools
  pip3 install numpy
  pip3 install sklearn
  pip3 install pandas
  pip3 install parse
  pip3 install matplotlib
  pip3 install IPython
  pip3 install ttkthemes
  

• Run VCBot

  python main.py
  

Work reference

Chame, H. F., Ahmadi, A., & Tani, J. (2020). A hybrid human-neurorobotics approach to primary intersubjectivity via active inference. Frontiers in Psychology, 11, 3207.

Contact

This program was implemented by: Hendry F. Chame

Lab Cognitive Neurorobotics Research Unit (CNRU)

Institution Okinawa Institute of Science and Technology Graduate University (OIST)

Address 1919-1, Tancha, Onna, Kunigami District, Okinawa 904-0495, Japan

E-mail [email protected]