-> To the work’s website.

-> May 2021
-> Keywords: Computational Art, Machine Learning, Cellular Automata, Chinese Characters
-> Documentation

New CharActers is a series of experiments that explore the potential when Neural Cellular Automata and Chinese characters meet. It involves researches and studies on machine learning, neural networks and the transformation of Chinese characters.

Built with Tensorflow framework, it trys to explore and reveal the interconnection between different Chinese character scripts - from ancient ages to modern times - with code as the medium.

The algorithm is largely based on the Growing Neural Cellular Automata model in Differentiable Self-organizing Systems, an experimental format collecting invited short articles delving into differentiable self-organizing systems, proposed by Alexander Mordvintsev, Ettore Randazzo, Eyvind Niklasson, and Michael Levin.

This neural cellular automata system is designed for imitating the same plasticity and robustness as biological life in the digital world: structures and machines could grow and repair themselves. It focuses on Cellular Automata models as a roadmap for the effort of identifying cell-level rules which give rise to complex, regenerative behavior of the collective.