Optimizing Metamaterial Inverse Design with 3D Conditional Diffusion Model and Data Augmentation (Adv. Mater. Technol. 14/2025)

Metamaterial Design In article number 2500293, Xiaoyang Zheng, Takayuki Yamada, and Junichiro Shiomi propose a novel 3D conditional diffusion model (3D-CDM), combined with a data augmentation technique to expand the small dataset, aimed at enhancing the inverse design of voxel-based metamaterials. The 3D-CDM generates realistic 3D metamaterials with targeted properties, outperforming existing methods in both fidelity and diversity.