Institutional Repository of Vilnius University
Ortografinių nuotraukų detalus lyginimas /
Abstract
dc:descriptionDrone position is usually determined using the Global Positioning system (GPS). However, GPS might not be always available for a variety of reasons. Due to that, it is useful to have alternative methods of determining a drone's position. One of the ways this can be achieved is by comparing pictures that the drone has taken to a built-in map. In this research paper, the effectiveness of the modern CenterNet2 neural network will be tested in the task of comparing images. The accuracy and speed of this network will be compared to other widely used networks - VGG16, ResNet50, MobileNet, EfficientNet and EfficientNetV2. These networks will be used as backbones in custom-made triplet networks.
Degree
thesis:*- Grantor dc:publisher
- Institutional Repository of Vilnius University
- Year dc:date
- 2024
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Vaiškūnas, Gabrielius,
Rights
dc:rights- Statement dc:rights
-
- info:eu-repo/semantics/openAccess
- Language dc:language
- lit
Identifiers
dc:identifier.*- Repository record dc:identifier
- https://repository.vu.lt/VU:ELABAETD210648193&prefLang=en_US