Inverse Design of Metamaterials with Manufacturing-Aware Spectrum-to-Shape Diffusion Models

The dataset includes detailed information on the MIM tri-layer metamaterial structures designed and used for training the DiffMeta framework. Specifically, it contains 60000 data:

Structural Data: Detailed geometric patterns and composition parameters of the designed MIM tri-layer metamaterial structures.

Spectral Data: Spectral measurements on MIM tri-layer metamaterial structures, including emissivity, reflectivity and transmittance spectra across a range of wavelengths.

The dataset is meticulously organized to facilitate the replication of our study and support further research in the field of metamaterial design.