Hyperparameter tuning of optical neural network classifiers for high-order Gaussian beams

High-order Gaussian beams with multiple propagation modes have been studied for free-space optical communications. Fast classification of beams using a diffractive deep neural network ((DNN)-N-2) has been proposed. (DNN)-N-2 optimization is important because it has numerous hyperparameters, such as interlayer distances and mode combinations. In this study, we classify Hermite-Gaussian beams, which are high-order Gaussian beams, using a (DNN)-N-2, and automatically tune one of its hyperparameters known as the interlayer distance. We used the tree-structured Parzen estimator, a hyperparameter auto-tuning algorithm, to search for the best model. As a result, the proposed method improved the classification accuracy in a 16 mode classification from 98.3% in the case of equal spacing of layers to 98.8%. In a 36 mode classification, the proposed method significantly improved the classification accuracy from 84.9% to 94.9%. In addition, we confirmed that accuracy by auto-tuning improves as the number of classification modes increases. (C) 2022 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement

Automated summary

This study proposes a method using a deep neural network for beam classification and improves the classification accuracy by automatically tuning the hyperparameters. The results demonstrate that the proposed method is effective in improving the accuracy, especially when there are more modes to classify.