TY - GEN
T1 - Learning a versatile representation of SAR data for regression and segmentation by leveraging self-supervised despeckling with MERLIN
AU - Dalsasso, Emanuele
AU - Rambour, Clément
AU - Denis, Loïc
AU - Tupin, Florence
N1 - Publisher Copyright:
© VDE VERLAG GMBH ∙ Berlin ∙ Offenbach.
PY - 2024/1/1
Y1 - 2024/1/1
N2 - Synthetic Aperture Radar (SAR) images are abundantly available, yet labels are often missing. Thus, training a neural network in a fully supervised manner is arduous. In this work, we leverage MERLIN, a self-supervised despeckling algorithm, to learn a mapping of SAR images into a representation space shared among despeckling, building segmentation and height regression. Our experiments demonstrate that the joint training of a neural network for these three tasks reduces considerably the need for labeled data to solve the supervised tasks: positive results are obtained even when only 1% of the dataset is annotated.
AB - Synthetic Aperture Radar (SAR) images are abundantly available, yet labels are often missing. Thus, training a neural network in a fully supervised manner is arduous. In this work, we leverage MERLIN, a self-supervised despeckling algorithm, to learn a mapping of SAR images into a representation space shared among despeckling, building segmentation and height regression. Our experiments demonstrate that the joint training of a neural network for these three tasks reduces considerably the need for labeled data to solve the supervised tasks: positive results are obtained even when only 1% of the dataset is annotated.
M3 - Conference contribution
AN - SCOPUS:85193942430
T3 - Proceedings of the European Conference on Synthetic Aperture Radar, EUSAR
SP - 1265
EP - 1270
BT - EUSAR 2024 - 15th European Conference on Synthetic Aperture Radar
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 15th European Conference on Synthetic Aperture Radar, EUSAR 2024
Y2 - 23 April 2024 through 26 April 2024
ER -