Semi-Supervised Learning for Ill-Posed Polarimetric SAR Classification
In recent years, the interest in semi-supervised learning has increased, combining supervised and unsupervised learning approaches. This is especially valid for classification applications in remote sensing, while the data acquisition rate in current systems has become fairly large considering high-...
Main Authors: | , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2014-05-01
|
Series: | Remote Sensing |
Subjects: | |
Online Access: | http://www.mdpi.com/2072-4292/6/6/4801 |
id |
doaj-90044838f5ac423f95b8fb25315a6d60 |
---|---|
record_format |
Article |
spelling |
doaj-90044838f5ac423f95b8fb25315a6d602020-11-24T22:17:05ZengMDPI AGRemote Sensing2072-42922014-05-01664801483010.3390/rs6064801rs6064801Semi-Supervised Learning for Ill-Posed Polarimetric SAR ClassificationStefan Uhlmann0Serkan Kiranyaz1Moncef Gabbouj2Department of Signal Processing, Tampere University of Technology, 33720 Tampere, FinlandDepartment of Signal Processing, Tampere University of Technology, 33720 Tampere, FinlandDepartment of Signal Processing, Tampere University of Technology, 33720 Tampere, FinlandIn recent years, the interest in semi-supervised learning has increased, combining supervised and unsupervised learning approaches. This is especially valid for classification applications in remote sensing, while the data acquisition rate in current systems has become fairly large considering high- and very-high resolution data; yet on the other hand, the process of obtaining the ground truth data may be cumbersome for such large repositories. In this paper, we investigate the application of semi-supervised learning approaches and particularly focus on the small sample size problem. To that extend, we consider two basic unsupervised approaches by enlarging the initial labeled training set as well as an ensemble-based self-training method. We propose different strategies within self-training on how to select more reliable candidates from the pool of unlabeled samples to speed-up the learning process and to improve the classification performance of the underlying classifier ensemble. We evaluate the effectiveness of the proposed semi-supervised learning approach over polarimetric SAR data. Results show that the proposed self-training approach using an ensemble-based classifier that is initially trained over a small training set can achieve a similar performance level of a fully supervised learning approach where the training is performed over significantly larger labeled data. Considering the difficulties of the manual data labeling in such massive volumes of SAR repositories, this is indeed a promising accomplishment for semi-supervised SAR classification.http://www.mdpi.com/2072-4292/6/6/4801semi-supervisedmachine learningensembleSARsuperpixel |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Stefan Uhlmann Serkan Kiranyaz Moncef Gabbouj |
spellingShingle |
Stefan Uhlmann Serkan Kiranyaz Moncef Gabbouj Semi-Supervised Learning for Ill-Posed Polarimetric SAR Classification Remote Sensing semi-supervised machine learning ensemble SAR superpixel |
author_facet |
Stefan Uhlmann Serkan Kiranyaz Moncef Gabbouj |
author_sort |
Stefan Uhlmann |
title |
Semi-Supervised Learning for Ill-Posed Polarimetric SAR Classification |
title_short |
Semi-Supervised Learning for Ill-Posed Polarimetric SAR Classification |
title_full |
Semi-Supervised Learning for Ill-Posed Polarimetric SAR Classification |
title_fullStr |
Semi-Supervised Learning for Ill-Posed Polarimetric SAR Classification |
title_full_unstemmed |
Semi-Supervised Learning for Ill-Posed Polarimetric SAR Classification |
title_sort |
semi-supervised learning for ill-posed polarimetric sar classification |
publisher |
MDPI AG |
series |
Remote Sensing |
issn |
2072-4292 |
publishDate |
2014-05-01 |
description |
In recent years, the interest in semi-supervised learning has increased, combining supervised and unsupervised learning approaches. This is especially valid for classification applications in remote sensing, while the data acquisition rate in current systems has become fairly large considering high- and very-high resolution data; yet on the other hand, the process of obtaining the ground truth data may be cumbersome for such large repositories. In this paper, we investigate the application of semi-supervised learning approaches and particularly focus on the small sample size problem. To that extend, we consider two basic unsupervised approaches by enlarging the initial labeled training set as well as an ensemble-based self-training method. We propose different strategies within self-training on how to select more reliable candidates from the pool of unlabeled samples to speed-up the learning process and to improve the classification performance of the underlying classifier ensemble. We evaluate the effectiveness of the proposed semi-supervised learning approach over polarimetric SAR data. Results show that the proposed self-training approach using an ensemble-based classifier that is initially trained over a small training set can achieve a similar performance level of a fully supervised learning approach where the training is performed over significantly larger labeled data. Considering the difficulties of the manual data labeling in such massive volumes of SAR repositories, this is indeed a promising accomplishment for semi-supervised SAR classification. |
topic |
semi-supervised machine learning ensemble SAR superpixel |
url |
http://www.mdpi.com/2072-4292/6/6/4801 |
work_keys_str_mv |
AT stefanuhlmann semisupervisedlearningforillposedpolarimetricsarclassification AT serkankiranyaz semisupervisedlearningforillposedpolarimetricsarclassification AT moncefgabbouj semisupervisedlearningforillposedpolarimetricsarclassification |
_version_ |
1725786707221020672 |