Predicting Chern numbers in photonic crystals using generative adversarial network-based data augmentation.

The Chern number is the core of topological photonics, which is used to describe the topological properties of photonic crystals and other optical systems to realize the functional transmission and the control of photons within materials. However, the calculation process of Chern numbers is complex and time-consuming. To address this issue, we use the deep learning accompanied with Maxwell's equations to predict the Chern number of a two-dimensional photonic crystal with a square lattice in this paper. We propose a numerical-to-image generative adversarial networks (GANs) augmentation method to solve the problem of insufficient training data. Our method demonstrates excellent predictive performance on the test dataset, achieving an average accuracy of 92.25%. Besides that, the proposed data augmentation method can significantly improve the accuracy of Chern number predictions by 7.95%, compared with the method that did not use this approach. This method offers what we believe to be a novel solution to the challenge of limited numerical data samples in deep learning applications like complex calculations of physical quantities. It may also have certain potential to improve deep learning algorithms in other fields of science and engineering.