Publications by authors named "Kazuma Matsuyama"
Entropy (Basel)
April 2023
Article Synopsis
- Generative design automates design processes but lacks the ability to assess psychological aspects like beauty, requiring designers to rely on their own judgment for shape selection.
- The study systematically classified curvature types and feature descriptors to better quantify "complexity" in 3D shapes, focusing on five curvatures and a feature descriptor derived from entropy.
- Findings indicated that using mean curvature provided the highest correlation with sensory evaluations of complexity, while Casorati curvature showed a strong signal-to-noise ratio, setting a foundation for improving generative design through psychological insights.
