TY - GEN
T1 - Exploring Data Quantity Requirements for Domain Adaptation in the Classification of Satellite Image Time Series
AU - Lucas, Benjamin
AU - Pelletier, Charlotte
AU - Inglada, Jordi
AU - Schmidt, Daniel
AU - Webb, Geoffrey I.
AU - Petitjean, Francois
N1 - Publisher Copyright:
© 2019 IEEE.
PY - 2019/8
Y1 - 2019/8
N2 - Land cover maps are a vital input variable in all types of environmental research and management. However the modern state-of-The-Art machine learning techniques used to create them require substantial training data to produce optimal accuracy. Domain Adaptation is one technique researchers might use when labelled training data are unavailable or scarce. This paper looks at the result of training a convolutional neural network model on a region where data are available (source domain), and then adapting this model to another region (target domain) by retraining it on the available labelled data, and in particular how these results change with increasing data availability. Our experiments performing domain adaptation on satellite image time series, draw three interesting conclusions: (1) a model trained only on data from the source domain delivers 73.0% test accuracy on the target domain; (2) when all of the weights are retrained on the target data, over 16,000 instances were required to improve upon the accuracy of the source-only model; and (3) even if sufficient data is available in the target domain, using a model pretrained on a source domain will result in better overall test accuracy compared to a model trained on target domain data only-88.9% versus 84.7%.
AB - Land cover maps are a vital input variable in all types of environmental research and management. However the modern state-of-The-Art machine learning techniques used to create them require substantial training data to produce optimal accuracy. Domain Adaptation is one technique researchers might use when labelled training data are unavailable or scarce. This paper looks at the result of training a convolutional neural network model on a region where data are available (source domain), and then adapting this model to another region (target domain) by retraining it on the available labelled data, and in particular how these results change with increasing data availability. Our experiments performing domain adaptation on satellite image time series, draw three interesting conclusions: (1) a model trained only on data from the source domain delivers 73.0% test accuracy on the target domain; (2) when all of the weights are retrained on the target data, over 16,000 instances were required to improve upon the accuracy of the source-only model; and (3) even if sufficient data is available in the target domain, using a model pretrained on a source domain will result in better overall test accuracy compared to a model trained on target domain data only-88.9% versus 84.7%.
UR - http://www.scopus.com/inward/record.url?scp=85074298390&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85074298390&partnerID=8YFLogxK
U2 - 10.1109/Multi-Temp.2019.8866898
DO - 10.1109/Multi-Temp.2019.8866898
M3 - Conference contribution
AN - SCOPUS:85074298390
T3 - 2019 10th International Workshop on the Analysis of Multitemporal Remote Sensing Images, MultiTemp 2019
BT - 2019 10th International Workshop on the Analysis of Multitemporal Remote Sensing Images, MultiTemp 2019
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 10th International Workshop on the Analysis of Multitemporal Remote Sensing Images, MultiTemp 2019
Y2 - 5 August 2019 through 7 August 2019
ER -