Institutional Repository of Vilnius University
Besikeičiančio žemės reljefo stebėjimas kompiuterine rega /
Abstract
dc:descriptionMonitoring the changing terrain is a pressing issue in the face of increasing climate change and human influences. This challenge can be addressed by using computer vision and deep learning predictive models to forecast future data. In this work, long short-term memory (LSTM) and transformer-based deep learning models are investigated and applied to predict the earth change observations using satellite imagery. These models are also combined to extract a more accurate prediction. The results show that the most accurate prediction in terms of mean absolute error is obtained by combining the LSTM model and the transformer, and the transformer's prediction is more accurate than that of the LSTM model.
Degree
thesis:*- Grantor dc:publisher
- Institutional Repository of Vilnius University
- Year dc:date
- 2023
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Kopūstas, Benas,
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:ELABAETD192846710&prefLang=en_US