In remote sensing image classification, it is commonly assumed that the distribution of the classes is stable over the entire image. This way, training pixels labeled by photointerpretation are assumed to be representative of the whole image. However, differences in distribution of the classes throughout the image make this assumption weak and a model built on a single area may be suboptimal when applied to the rest of the image. In this paper, we investigate the use of active learning to correct the shifts that may appear when training and test data do not come from the same distribution. Experiments are carried out on a VHR remote sensing classification scenario showing that active learning can effectively learn the covariance shift and provide robust solutions. © 2011 IEEE.
Dataset shift adaptation with active queries / Tuia, D.; Pasolli, E.; Emery, W. J.. - (2011), pp. 121-124. (Intervento presentato al convegno IEEE GRSS and ISPRS Joint Urban Remote Sensing Event, JURSE 2011 tenutosi a Munich, deu nel 2011) [10.1109/JURSE.2011.5764734].
Dataset shift adaptation with active queries
Pasolli E.;
2011
Abstract
In remote sensing image classification, it is commonly assumed that the distribution of the classes is stable over the entire image. This way, training pixels labeled by photointerpretation are assumed to be representative of the whole image. However, differences in distribution of the classes throughout the image make this assumption weak and a model built on a single area may be suboptimal when applied to the rest of the image. In this paper, we investigate the use of active learning to correct the shifts that may appear when training and test data do not come from the same distribution. Experiments are carried out on a VHR remote sensing classification scenario showing that active learning can effectively learn the covariance shift and provide robust solutions. © 2011 IEEE.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
