Análise de modelos adversariais aplicados à classificação de sinais de Sonar Passivo

Julio De Castro Vargas Fernandes, José de Seixas, Natanael Nunes de Moura Junior

DOI: 10.14209/sbrt.2019.1570558846

Evento: XXXVII Simpósio Brasileiro de Telecomunicações e Processamento de Sinais (SBrT2019)

Keywords:

Abstract

In naval warfare, several techniques have been developed for the detection and classification of ships. Given the confidential nature of the data it is extremely difficult to get a hold of large quantities of data which makes it extremely hard to use techniques that rely on abundant data, like deep learning. This paper proposes the use of generative adversarial neural networks for the classification of passive sonar signals. Since the proposed technique trains an synthetic image generator, which will later be used for the training of a classifier, its quality will be evaluated. The proposed method achieved an efficiency of 99,2 +- 1,0.

Download