Machine Learning for artists / 3Deeprinting
This project has been developed during the one-week workshop “Machine Learning for artists” held by Gene Kogan in Opendot on November 2016, here you can find the online repository > https://github.com/opendot/ml4a-3Deeprinting
/ Concept
3D models generator based on Recurrent Neural Network (LSTM). For more details see The Unreasonable Effectiveness of Recurrent Neural Networks blog post. Generate new 3D models starting from already published ones. We didn’t really know what we were doing, and that was awesome.
/ Realisation
Main steps: collect data scraping of STL models about pots and vases convert STL from/to GCode and Raw generate models, after training the generative model.
/ Team
Sabina Barcucci, Patrizia Bolzan, Gene Kogan, Massimo Santini, Piero Savastano, Alessandro Zomparelli and Opendot