TY - GEN
T1 - CNN Sensitivity Analysis for Land Cover Map Models Using Sparse and Heterogeneous Satellite Data
AU - Moreno, Sebastián
AU - Lopatin, Javier
AU - Corvalán, Diego
AU - Bravo-Diaz, Alejandra
N1 - Publisher Copyright:
© The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.
PY - 2025
Y1 - 2025
N2 - Land cover maps provide detailed information on the land use of territories, which is useful for public policy making. Constant changes in the landscape limit the usefulness of these maps over time, so they need to be constantly updated. In this context, remote sensing images combined with the use of deep neural networks can be used for this purpose. Although several models are trained on different datasets, we do not know their ability to transfer the learned patterns to new data. In this paper, we evaluate several pre-trained semantic segmentation models on deep convolutional neural networks (CNN) using freely available global RGB data from Sentinel-2. Four CNN models with 32 different architectures were evaluated on data from three continents, on seven different classes. The results show that the best model is the PSPNet with seresnet18, obtaining a test macro F1 score of 0.4950 when the model is trained with data augmentation and fine-tuning.
AB - Land cover maps provide detailed information on the land use of territories, which is useful for public policy making. Constant changes in the landscape limit the usefulness of these maps over time, so they need to be constantly updated. In this context, remote sensing images combined with the use of deep neural networks can be used for this purpose. Although several models are trained on different datasets, we do not know their ability to transfer the learned patterns to new data. In this paper, we evaluate several pre-trained semantic segmentation models on deep convolutional neural networks (CNN) using freely available global RGB data from Sentinel-2. Four CNN models with 32 different architectures were evaluated on data from three continents, on seven different classes. The results show that the best model is the PSPNet with seresnet18, obtaining a test macro F1 score of 0.4950 when the model is trained with data augmentation and fine-tuning.
KW - Land cover maps
KW - deep neural networks
KW - transfer learning
UR - http://www.scopus.com/inward/record.url?scp=85210227499&partnerID=8YFLogxK
U2 - 10.1007/978-3-031-76607-7_5
DO - 10.1007/978-3-031-76607-7_5
M3 - Conference contribution
AN - SCOPUS:85210227499
SN - 9783031766060
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 63
EP - 77
BT - Progress in Pattern Recognition, Image Analysis, Computer Vision, and Applications - 27th Iberoamerican Congress, CIARP 2024, Proceedings
A2 - Hernández-García, Ruber
A2 - Barrientos, Ricardo J.
A2 - Velastin, Sergio A.
PB - Springer Science and Business Media Deutschland GmbH
T2 - 27th Iberoamerican Congress on Progress in Pattern Recognition, Image Analysis, Computer Vision, and Applications, CIARP 2024
Y2 - 26 November 2024 through 29 November 2024
ER -
