Skip to content
Snippets Groups Projects

VS2N

Last edited by Hammouda Elbez

Description

VS2N 1 (Visualization tool for Spiking Neural Networks) is an interactive web-based tool. It is designed to analyze and visualize the activity we collect from SNN simulators.

We can use VS2N to:

  • Follow the evolution of our network during training.
  • Validate hypotheses.
  • Debug any newly implemented component.

VS2N is a module-based tool written in Python, those modules concern defined questions or elements to visualize, and it is possible to add new modules. Each module contains different visualizations.

Initial requirements for contributions :

  • Python (Flask)
  • Dash Plotly (to implement different visualizations)
  • Apache Spark (for any heavy pre-calculations if needed)
  • MongoDB (to store and read data)

VS2N

Installation

1- Install MongoDB

Use the following link to install MongoDB.

2- Install Apache Spark and other dependencies

Execute this command to install the needed libraries

pip install -r requirements.txt

Execution

  1. If you want to add authentication to VS2N (optional):
  • Generate users credentials (Email/Password) using python manageUsers.py

  1. Rename default.config.py to config.py

  2. If a password protects MongoDB:

  • Add MongoDB credentials in config.py
  1. Using Python >= 3.0 (make sure MongoDB is running), run python VS2N.py to start VS2N. If everything is fine, the web interface should appear automatically.

Docker image

You can also use a docker image to run VS2N : https://hub.docker.com/r/helbez/vs2n

Official Documentation

VS2N Documentation:
https://gitlab.univ-lille.fr/bioinsp/VS2N/-/wikis/home

References

  1. H. Elbez, M. K. Benhaoua, P. Devienne and P. Boulet, "VS2N : Interactive Dynamic Visualization and Analysis Tool for Spiking Neural Networks," 2021 International Conference on Content-Based Multimedia Indexing (CBMI), pp. 1-6, 2021 ↩

Comments

    Please register or sign in to add a comment.