TY  - JOUR
AU  - Antonijević, Ognjen
AU  - Jelić, Slobodan
AU  - Bajat, Branislav
AU  - Kilibarda, Milan
PY  - 2023
UR  - https://grafar.grf.bg.ac.rs/handle/123456789/3098
AB  - This paper presents a transfer learning approach to the crop classification problem based on time series of images from the Sentinel-2 dataset labeled for two regions: Brittany (France) and Vojvodina (Serbia). During preprocessing, cloudy images are removed from the input data, the time series are interpolated over the time dimension, and additional remote sensing indices are calculated. We chose TransformerEncoder as the base model for knowledge transfer from source to target domain with French and Serbian data, respectively. Even more, the accuracy of the base model with the preprocessing step is improved by 2% when trained and evaluated on the French dataset. The transfer learning approach with fine-tuning of the pre-trained weights on the French dataset outperformed all other methods in terms of overall accuracy 0.94 and mean class recall 0.907 on the Serbian dataset. Our partially fine-tuned model improved recall of crop types that were poorly classified by the base model. In the case of sugar beet, class recall is improved by 85.71%.
PB  - Springer
T2  - Journal of Big Data
T1  - Transfer learning approach based on satellite image time series for the crop classification problem
VL  - 10
DO  - 10.1186/s40537-023-00735-2
ER  - 

@article{
author = "Antonijević, Ognjen and Jelić, Slobodan and Bajat, Branislav and Kilibarda, Milan",
year = "2023",
abstract = "This paper presents a transfer learning approach to the crop classification problem based on time series of images from the Sentinel-2 dataset labeled for two regions: Brittany (France) and Vojvodina (Serbia). During preprocessing, cloudy images are removed from the input data, the time series are interpolated over the time dimension, and additional remote sensing indices are calculated. We chose TransformerEncoder as the base model for knowledge transfer from source to target domain with French and Serbian data, respectively. Even more, the accuracy of the base model with the preprocessing step is improved by 2% when trained and evaluated on the French dataset. The transfer learning approach with fine-tuning of the pre-trained weights on the French dataset outperformed all other methods in terms of overall accuracy 0.94 and mean class recall 0.907 on the Serbian dataset. Our partially fine-tuned model improved recall of crop types that were poorly classified by the base model. In the case of sugar beet, class recall is improved by 85.71%.",
publisher = "Springer",
journal = "Journal of Big Data",
title = "Transfer learning approach based on satellite image time series for the crop classification problem",
volume = "10",
doi = "10.1186/s40537-023-00735-2"
}

Antonijević, O., Jelić, S., Bajat, B.,& Kilibarda, M.. (2023). Transfer learning approach based on satellite image time series for the crop classification problem. in Journal of Big Data
Springer., 10.
https://doi.org/10.1186/s40537-023-00735-2

Antonijević O, Jelić S, Bajat B, Kilibarda M. Transfer learning approach based on satellite image time series for the crop classification problem. in Journal of Big Data. 2023;10.
doi:10.1186/s40537-023-00735-2 .

Antonijević, Ognjen, Jelić, Slobodan, Bajat, Branislav, Kilibarda, Milan, "Transfer learning approach based on satellite image time series for the crop classification problem" in Journal of Big Data, 10 (2023),
https://doi.org/10.1186/s40537-023-00735-2 . .