TY - GEN
T1 - A Weakly-Supervised, Multitask Deep Learning Framework for Shadow Mitigation in Remote Sensing Imagery
AU - Couwenhoven, Scott D.
AU - Ientilucci, Emmett J.
AU - Park, Byung H.
AU - Hughes, David
N1 - Publisher Copyright:
© 2022 IEEE.
PY - 2022
Y1 - 2022
N2 - We propose a weakly-supervised, multitask framework for training a convolutional neural network to solve the problem of cloud shadow mitigation given only cloud and shadow masks as labels. The network minimizes the Wasserstein distance between shadows and their proximal sunlit neighborhoods, generating a supervisory signal directly from within the input image. We extract further utility from the shadow mask through multitask learning by introducing an auxiliary task of shadow segmentation. Our approach is advantageous since it performs mitigation in an end-to-end framework which requires only a shadowed image for inference. We apply this process to the Landsat 8 OLI SPARCS validation data set and demonstrate plausible results.
AB - We propose a weakly-supervised, multitask framework for training a convolutional neural network to solve the problem of cloud shadow mitigation given only cloud and shadow masks as labels. The network minimizes the Wasserstein distance between shadows and their proximal sunlit neighborhoods, generating a supervisory signal directly from within the input image. We extract further utility from the shadow mask through multitask learning by introducing an auxiliary task of shadow segmentation. Our approach is advantageous since it performs mitigation in an end-to-end framework which requires only a shadowed image for inference. We apply this process to the Landsat 8 OLI SPARCS validation data set and demonstrate plausible results.
KW - cloud shadow mitigation
KW - deep learning
KW - multitask learning
KW - satellite imagery
KW - weak supervision
UR - http://www.scopus.com/inward/record.url?scp=85140407073&partnerID=8YFLogxK
U2 - 10.1109/IGARSS46834.2022.9883550
DO - 10.1109/IGARSS46834.2022.9883550
M3 - Conference contribution
AN - SCOPUS:85140407073
T3 - International Geoscience and Remote Sensing Symposium (IGARSS)
SP - 619
EP - 622
BT - IGARSS 2022 - 2022 IEEE International Geoscience and Remote Sensing Symposium
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 2022 IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2022
Y2 - 17 July 2022 through 22 July 2022
ER -