Research on UAV Target Recognition Technology Based on Federated Learning
- Resource Type
- Conference
- Authors
- Fang, Lu-lu; Hu, Hai-rong; Pu, Wei; Bi, Jian-quan
- Source
- 2021 2nd International Conference on Computer Engineering and Intelligent Control (ICCEIC) ICCEIC Computer Engineering and Intelligent Control (ICCEIC), 2021 2nd International Conference on. :119-122 Nov, 2021
- Subject
- Computing and Processing
Training
Data privacy
Target recognition
Computational modeling
Neural networks
Transfer learning
Computer architecture
Federated Learning
UAV
target recognition
YOLOv3-Tiny
- Language
This paper analyzes the difficulty of data sharing in the current UAV target recognition model training, and designs a UAV target recognition architecture based on federated learning. And combined with the advantages of YOLOv3 to improve the YOLOv3-Tiny model. Finally, the model is trained in two different scenarios through simulation experiments. The results show that the performance of the model is improved on the basis of protecting data privacy. It provides an index reference for the next step in the in-service assessment of UAV equipment target recognition capabilities.