Artificial Intelligence Simulates Tree Growth and Shape, Revolutionizing Digital Tree Models

Researchers from Purdue University, in collaboration with Kiel University in Germany, have developed artificial intelligence (AI) models that can simulate tree growth and shape. Inspired by DNA, the AI models compress the information needed for encoding tree form into a megabyte-sized neural model. These AI models generate detailed geometry of complex tree models, which are crucial for architecture, urban planning, gaming, and entertainment industries. The traditional complex simulation algorithms for creating digital tree models are now being replaced by AI-based models that compress complex behavior into a small amount of data. This advancement is expected to significantly improve existing methods for digital tree models, covering various tree species. With the use of AI, the team aims to overcome the deficiency in training data and decouple a tree's intrinsic properties from its environmental response.