Configurator help    1.1.20201205
Generative algorithms are able to create awesome and beautiful artworks. However those algorithms usually depend on a high number of parameters which meanings are not easy to interpret. The GENEVL configurator allows you to explore all the universe of parameter combinations handling a simple UI control and regardless of the algorithm complexity. Additionally it includes several tools for saving and sharing the explored configurations as well as utilities for obtaining the generated models.

How it works

Use the evolution control bar that directly sets the evolution speed of the model.

Touch the bar within the positive half (∞) in order to evolve the model seeking for characteristics you like. If you don't like a change then simply touch the bar within the negative half (○) so the application involves the model undoing the evolution history. The next time you evolve the model the tool will generate a different evolutionary path. Sometimes you have to pass through configurations you don't like to reach one you like.
Playing this game of evolution-involution teaches the internal AI engine of the application how to better generate new evolutionary paths for accelerating the convergence. Convergence is reached when the user is satisfied with the model configuration and no more evolution is necessary.

Other functions

Press the button  to access the application menu:
  • Evolve geometry: Disable this option to preserve geometry while evolving other features.
  • Evolve colors: Disable this option to preserve colors while evolving other features.
  • Save configuration: Saves the current configuration to the local drive as a TXT file. Neither the evolution history nor the training of the internal AI engine are saved.
  • Load configuration: Loads a configuration from the local drive previously saved as a TXT file. Once the file is chosen the application displays a dialog offering import options. This possibility is very useful for combining characteristics from different model configurations. When a configuration is loaded into the tool it becomes the starting point for the next evolution process and any previous evolution history is lost.
  • Save snapshot: Saves to the local drive a PNG image of the model with transparent background.
  • Share model: Stores the current configuration on the server and shows the URL and social buttons for sharing it.