Active Learning Methods for Remote Sensing Image Classification

Tuia, Devis; Ratle, Frédéric; Pacifici, Fabio; Kanevski, Mikhaïl; Emery, William J.

doi:10.1109/tgrs.2008.2010404

Cited by 458 publications

(224 citation statements)

References 39 publications

Supporting

Mentioning

223

Contrasting

Order By: Relevance

“…In real applications with noisy data, sampling data with the highest inconsistency value measured by can result in noise or outliers being introduced into the training set. Thus, we introduce a relaxation variable , and first query samples from the current unlabeled data pool according to (4), and then further randomly select samples from this subset. The design of the regularizer is key to the success of the active learning strategy.…”

Section: Data Regularization Based Active Learingmentioning

confidence: 99%

“…Methods employ different query strategies, such as margin sampling [2], [4], [6], uncertainty sampling [3], [4], cost sensitive sampling [5], and the query-by-committee (QBC)-based method [4], [7]. The key is to select samples with higher uncertainty or which cause greater ambiguity for the classifier.…”

mentioning

confidence: 99%

“…The query-by-committee based method, which is a popular strategy in AL, utilizes the lack of consensus between a group of diverse classifier committee members [4], [7], whereby samples for which there is greatest disagreement among the committee are selected. For a discriminative classifier, another effective strategy is to select boundary samples, which often cause confusion to the classifier and are also key to building the classification hyperplane.…”

mentioning

confidence: 99%

“…For a discriminative classifier, another effective strategy is to select boundary samples, which often cause confusion to the classifier and are also key to building the classification hyperplane. A straightforward way for identifying boundary samples is to search for samples that are close to the current classification hyperplane [2], [4], [6], [8], e.g., margin sampling. Other strategies can also be applied, such as querying "nearby" samples that have greater inconsistency relative to the given sample by investigating local variation (such as using the local Laplacian graph [6]).…”

mentioning

confidence: 99%

See 3 more Smart Citations

Active Learning via Multi-View and Local Proximity Co-Regularization for Hyperspectral Image Classification

Wei

Crawford

2011

IEEE J. Sel. Top. Signal Process.

View full text Add to dashboard Cite

Abstract-A novel co-regularization framework for active learning is proposed for hyperspectral image classification. The first regularizer explores the intrinsic multi-view information embedded in the hyperspectral data. By adaptively and quantitatively measuring the disagreement level, it focuses only on samples with high uncertainty and builds a contention pool which is a small subset of the overall unlabeled data pool, thereby mitigating the computational cost. The second regularizer is based on the "consistency assumption" and designed on a spatial or the spectral based manifold space. It serves to further focus on the most informative samples within the contention pool by penalizing rapid changes in the classification function evaluated on proximally close samples in a local region. Such changes may be due to the lack of capability of the current learner to describe the unlabeled data. Incorporating manifold learning into the active learning process enforces the clustering assumption and avoids the degradation of the distance measure associated with the original high-dimensional spectral features. One spatial and two local spectral embedding methods are considered in this study, in conjunction with the support vector machine (SVM) classifier implemented with a radial basis function (RBF) kernel. Experiments show excellent performance on AVIRIS and Hyperion hyperspectral data as compared to random sampling and the state-of-the-art SVM SIMPLE .

show abstract

Section: Data Regularization Based Active Learingmentioning

confidence: 99%

mentioning

confidence: 99%

mentioning

confidence: 99%

mentioning

confidence: 99%

See 2 more Smart Citations

Active Learning via Multi-View and Local Proximity Co-Regularization for Hyperspectral Image Classification

Wei

Crawford

2011

IEEE J. Sel. Top. Signal Process.

View full text Add to dashboard Cite

show abstract

“…Active learning methods have demonstrated value where large training datasets are available, such as in remote sensing [17], text classification [18,19] and object recognition applications. However, medical image classification models are typically constructed from a limited pool of expert-chosen regions of interest (ROIs).…”

Section: Introductionmentioning

confidence: 99%

Active Relearning for Robust Supervised Training of Emphysema Patterns

et al. 2014

View full text Add to dashboard Cite

Radiologists are adept at recognizing the character and extent of lung parenchymal abnormalities in computed tomography (CT) scans. However, the inconsistent differential diagnosis due to subjective aggregation necessitates the exploration of automated classification based on supervised or unsupervised learning. The robustness of supervised learning depends on the training samples. Towards optimizing emphysema classification, we introduce a physician-in-the-loop feedback approach to minimize ambiguity in the selected training samples. An experienced thoracic radiologist selected 412 regions of interest (ROIs) across 15 datasets to represent 124, 129, 139 and 20 training samples of mild, moderate, severe emphysema and normal appearance, respectively. Using multi-view (multiple metrics to capture complementary features) inductive learning, an ensemble of seven unoptimized support vector models (SVM) each based on a specific metric was constructed in less than 6 s. The training samples were classified using seven SVM models and consensus labels were created using majority voting. In the active relearning phase, the ensemble-expert label conflicts were resolved by the expert. The efficacy and generality of active relearning feedback was assessed in the optimized parameter space of six general purpose classifiers across the seven dissimilarity metrics. The proposed just-in-time active relearning feedback with un-optimized SVMs yielded 15 % increase in classification accuracy and 25 % reduction in the number of support vectors. The average improvement in accuracy of six classifiers in their optimized parameter space was 21 %. The proposed cooperative feedback method enhances the quality of training samples used to construct automated classification of emphysematous CT scans. Such an approach could lead to substantial improvement in quantification of emphysema.

show abstract