The Synthetic Data Application in the UAV Recognition Systems Development
2023 IEEE 10th Jubilee Workshop on Advances in Information, Electronic and Electrical Engineering (AIEEE 2023): Proceedings 2023
Diāna Dupļevska, Vladislavs Medvedevs, Daniils Surmačs, Artūrs Āboltiņš

The increasing popularity and accessibility of un-manned aerial vehicles (UAVs) presents both opportunities and challenges. On the one hand, UAVs has a wide range of civilian, industrial, and military applications. On the other hand, the popularity of UAVs can lead to illegal or dangerous usage. Thus, the development of UAV recognition systems is crucial for ensuring safety and security. However, collecting and labeling large amounts of real-world data for training these systems can be time-consuming and labor-intensive.In this study, we propose a methodology, which can help to accelerate the development of new UAV recognition systems. This work demonstrates the effectiveness of training a neural network using a combination of real-world and synthetic data that can achieve similar performance to a network trained on real-world data only.


Keywords
Neural networks, Convolutional neural networks, Artificial neural networks, Synthetic data, Generative adversarial networks
DOI
10.1109/AIEEE58915.2023.10134962
Hyperlink
https://ieeexplore.ieee.org/document/10134962

Dupļevska, D., Medvedevs, V., Surmačs, D., Āboltiņš, A. The Synthetic Data Application in the UAV Recognition Systems Development. In: 2023 IEEE 10th Jubilee Workshop on Advances in Information, Electronic and Electrical Engineering (AIEEE 2023): Proceedings, Lithuania, Vilnius, 27-29 April, 2023. Piscataway: IEEE, 2023, pp.1-6. ISBN 979-8-3503-1179-2. e-ISBN 979-8-3503-1178-5. ISSN 2689-7334. e-ISSN 2689-7342. Available from: doi:10.1109/AIEEE58915.2023.10134962

Publication language
English (en)
The Scientific Library of the Riga Technical University.
E-mail: uzzinas@rtu.lv; Phone: +371 28399196