The Experts below are selected from a list of 17769 Experts worldwide ranked by ideXlab platform
John P. Caspersen - One of the best experts on this subject based on the ideXlab platform.
-
region merging using local spectral angle thresholds a more accurate method for hybrid segmentation of remote sensing images
Remote Sensing of Environment, 2017Co-Authors: Jian Yang, John P. CaspersenAbstract:Abstract Image segmentation is the decisive process in object-based image analysis, but segmenting a landscape scene into meaningful geo-objects remains a challenge. In recent years, there has been growing interest in hybrid methods that combine initial segmentation with subsequent region merging, because they exploit spectral signals across an entire geo-object, including both the boundary signals used to delineate the initial segments, and the interior signals used to merge similar segments. However, existing algorithms commonly use a single, global parameter to control the process of region merging, thereby limiting the goodness-of-fit between segments and geo-objects, since homogeneous and heterogeneous segments are treated equally. To overcome this limitation, we developed a new hybrid segmentation method that employs local spectral angle (SA) thresholds for region merging. We implemented our local SA method in three very different landscapes, then compared our region merging method to the global SA method, as well as the global elevation method used in System for Automated Geoscientific Analyses (SAGA). In all three landscapes, the results revealed that the local SA segmentation provides a better fit to reference polygons than the two global threshold methods, as Measured using a modified Discrepancy Measure for the purpose of geo-object recognition ( QR M ). We also found that the local SA method produced segments with a greater variation in size, indicating the method is effective for achieving multi-scale segmentation.
-
A Discrepancy Measure for segmentation evaluation from the perspective of object recognition
ISPRS Journal of Photogrammetry and Remote Sensing, 2015Co-Authors: Jian Yang, John P. Caspersen, Trevor A. JonesAbstract:Abstract Within the framework of geographic object-based image analysis (GEOBIA), segmentation evaluation is one of the most important components and thus plays a critical role in controlling the quality of GEOBIA workflow. Among a variety of segmentation evaluation methods and criteria, Discrepancy Measurement is believed to be the most useful and is therefore one of the most commonly employed techniques in many applications. Existing Measures have largely ignored the importance of object recognition in segmentation evaluation. In this study, a new Discrepancy Measure of segmentation evaluation index (SEI) redefines the corresponding segment using a two-sided 50% overlap instead of one-sided 50% overlap that has been commonly used. The effectiveness of SEI is further investigated using the schematic segmentation cases and remote sensing images. Results demonstrate that the proposed SEI outperforms the other two existing Discrepancy Measures, Euclidean Distance 2 (ED2) and Euclidean Distance 3 (ED3), both in terms of object recognition accuracy and identification of detailed segmentation differences.
Bernhard Scholkopf - One of the best experts on this subject based on the ideXlab platform.
-
domain adaptation with conditional transferable components
International Conference on Machine Learning, 2016Co-Authors: Mingming Gong, Tongliang Liu, Dacheng Tao, Clark Glymour, Bernhard ScholkopfAbstract:Domain adaptation arises in supervised learning when the training (source domain) and test (target domain) data have different distributions. Let X and Y denote the features and target, respectively, previous work on domain adaptation mainly considers the covariate shift situation where the distribution of the features P(X) changes across domains while the conditional distribution P(Y|X) stays the same. To reduce domain Discrepancy, recent methods try to find invariant components τ (X) that have similar P(τ (X)) on different domains by explicitly minimizing a distribution Discrepancy Measure. However, it is not clear if P(Y|τ (X)) in different domains is also similar when P(Y|X) changes. Furthermore, transferable components do not necessarily have to be invariant. If the change in some components is identifiable, we can make use of such components for prediction in the target domain. In this paper, we focus on the case where P(X|Y) and P(Y) both change in a causal system in which Y is the cause for X. Under appropriate assumptions, we aim to extract conditional transferable components whose conditional distribution P(τ (X)|Y) is invariant after proper location-scale (LS) transformations, and identify how P(Y) changes between domains simultaneously. We provide theoretical analysis and empirical evaluation on both synthetic and real-world data to show the effectiveness of our method.
Masashi Sugiyama - One of the best experts on this subject based on the ideXlab platform.
-
AAAI - Unsupervised Domain Adaptation Based on Source-Guided Discrepancy
Proceedings of the AAAI Conference on Artificial Intelligence, 2019Co-Authors: Seiichi Kuroki, Nontawat Charoenphakdee, Han Bao, Junya Honda, Issei Sato, Masashi SugiyamaAbstract:Unsupervised domain adaptation is the problem setting where data generating distributions in the source and target domains are different and labels in the target domain are unavailable. An important question in unsupervised domain adaptation is how to Measure the difference between the source and target domains. Existing Discrepancy Measures for unsupervised domain adaptation either require high computation costs or have no theoretical guarantee. To mitigate these problems, this paper proposes a novel Discrepancy Measure called source-guided Discrepancy (S-disc), which exploits labels in the source domain unlike the existing ones. As a consequence, S-disc can be computed efficiently with a finitesample convergence guarantee. In addition, it is shown that S-disc can provide a tighter generalization error bound than the one based on an existing Discrepancy Measure. Finally, experimental results demonstrate the advantages of S-disc over the existing Discrepancy Measures.
-
Domain Discrepancy Measure Using Complex Models in Unsupervised Domain Adaptation.
arXiv: Machine Learning, 2019Co-Authors: Jongyeong Lee, Nontawat Charoenphakdee, Seiichi Kuroki, Masashi SugiyamaAbstract:Appropriately evaluating the Discrepancy between domains is essential for the success of unsupervised domain adaptation. In this paper, we first point out that existing Discrepancy Measures are less informative when complex models such as deep neural networks are used, in addition to the facts that they can be computationally highly demanding and their range of applications is limited only to binary classification. We then propose a novel domain Discrepancy Measure, called the paired hypotheses Discrepancy (PHD), to overcome these shortcomings. PHD is computationally efficient and applicable to multi-class classification. Through generalization error bound analysis, we theoretically show that PHD is effective even for complex models. Finally, we demonstrate the practical usefulness of PHD through experiments.
Jian Yang - One of the best experts on this subject based on the ideXlab platform.
-
region merging using local spectral angle thresholds a more accurate method for hybrid segmentation of remote sensing images
Remote Sensing of Environment, 2017Co-Authors: Jian Yang, John P. CaspersenAbstract:Abstract Image segmentation is the decisive process in object-based image analysis, but segmenting a landscape scene into meaningful geo-objects remains a challenge. In recent years, there has been growing interest in hybrid methods that combine initial segmentation with subsequent region merging, because they exploit spectral signals across an entire geo-object, including both the boundary signals used to delineate the initial segments, and the interior signals used to merge similar segments. However, existing algorithms commonly use a single, global parameter to control the process of region merging, thereby limiting the goodness-of-fit between segments and geo-objects, since homogeneous and heterogeneous segments are treated equally. To overcome this limitation, we developed a new hybrid segmentation method that employs local spectral angle (SA) thresholds for region merging. We implemented our local SA method in three very different landscapes, then compared our region merging method to the global SA method, as well as the global elevation method used in System for Automated Geoscientific Analyses (SAGA). In all three landscapes, the results revealed that the local SA segmentation provides a better fit to reference polygons than the two global threshold methods, as Measured using a modified Discrepancy Measure for the purpose of geo-object recognition ( QR M ). We also found that the local SA method produced segments with a greater variation in size, indicating the method is effective for achieving multi-scale segmentation.
-
A Discrepancy Measure for segmentation evaluation from the perspective of object recognition
ISPRS Journal of Photogrammetry and Remote Sensing, 2015Co-Authors: Jian Yang, John P. Caspersen, Trevor A. JonesAbstract:Abstract Within the framework of geographic object-based image analysis (GEOBIA), segmentation evaluation is one of the most important components and thus plays a critical role in controlling the quality of GEOBIA workflow. Among a variety of segmentation evaluation methods and criteria, Discrepancy Measurement is believed to be the most useful and is therefore one of the most commonly employed techniques in many applications. Existing Measures have largely ignored the importance of object recognition in segmentation evaluation. In this study, a new Discrepancy Measure of segmentation evaluation index (SEI) redefines the corresponding segment using a two-sided 50% overlap instead of one-sided 50% overlap that has been commonly used. The effectiveness of SEI is further investigated using the schematic segmentation cases and remote sensing images. Results demonstrate that the proposed SEI outperforms the other two existing Discrepancy Measures, Euclidean Distance 2 (ED2) and Euclidean Distance 3 (ED3), both in terms of object recognition accuracy and identification of detailed segmentation differences.
Mingming Gong - One of the best experts on this subject based on the ideXlab platform.
-
domain adaptation with conditional transferable components
International Conference on Machine Learning, 2016Co-Authors: Mingming Gong, Tongliang Liu, Dacheng Tao, Clark Glymour, Bernhard ScholkopfAbstract:Domain adaptation arises in supervised learning when the training (source domain) and test (target domain) data have different distributions. Let X and Y denote the features and target, respectively, previous work on domain adaptation mainly considers the covariate shift situation where the distribution of the features P(X) changes across domains while the conditional distribution P(Y|X) stays the same. To reduce domain Discrepancy, recent methods try to find invariant components τ (X) that have similar P(τ (X)) on different domains by explicitly minimizing a distribution Discrepancy Measure. However, it is not clear if P(Y|τ (X)) in different domains is also similar when P(Y|X) changes. Furthermore, transferable components do not necessarily have to be invariant. If the change in some components is identifiable, we can make use of such components for prediction in the target domain. In this paper, we focus on the case where P(X|Y) and P(Y) both change in a causal system in which Y is the cause for X. Under appropriate assumptions, we aim to extract conditional transferable components whose conditional distribution P(τ (X)|Y) is invariant after proper location-scale (LS) transformations, and identify how P(Y) changes between domains simultaneously. We provide theoretical analysis and empirical evaluation on both synthetic and real-world data to show the effectiveness of our method.