An abstract representation of integrating inverse design on metamaterials using deep learning architectures.

• The initial task was to simulate multiple simulations having different parameters using CST Studio Suite to generate resonance graphs of the particular metamaterial.
• Using these graphs we need to understand the resonating point
• Example

• With the help of several graphs obtained from simulating numerous parameters on metamaterials, we extracted the information present and collectively stored in the dataset.
• These datasets were used as a base to train the models for inverse design.
• Sample Dataset

• We created a couple of machine learning models using these datasets.

• Implementation of a deep learning model was done to make better predictions.
• 1D CNN layers were used to extract information from the graphs and make predictions of the parameters used.

• Basic frontend implementation of model usage was created, to showcase the working of the model.
• Implementation