Classification of Solar Radio Spectrum Based on Swin Transformer

TL;DR

This paper introduces a Swin transformer-based method for classifying solar radio spectrums, achieving high accuracy and significantly reducing model size compared to traditional CNNs, aiding real-time space weather monitoring.

Contribution

It proposes a transfer learning approach using Swin transformer for solar radio spectrum classification, with improved accuracy and reduced model complexity.

Findings

Achieved 100% true positive rate in classification.

Model parameters are reduced by 80% compared to VGG16.

Outperforms previous methods in accuracy.

Abstract

Solar radio observation is a method used to study the Sun. It is very important for space weather early warning and solar physics research to automatically classify solar radio spectrums in real time and judge whether there is a solar radio burst. As the number of solar radio burst spectrums is small and uneven, this paper proposes a classification method for solar radio spectrums based on the Swin transformer. First, the method transfers the parameters of the pretrained model to the Swin transformer model. Then, the hidden layer weights of the Swin transformer are frozen, and the fully connected layer of the Swin transformer is trained on the target dataset. Finally, pa-rameter tuning is performed. The experimental results show that the method can achieve a true positive rate of 100%, which is more accurate than previous methods. Moreover, the number of our model parameters is only 20 million, which is 80% lower than that of the traditional VGG16 con-volutional neural network with more than 130 million parameters.