- 标题
- 摘要
- 关键词
- 实验方案
- 产品
-
Material Decomposition in X-ray Spectral CT Using Multiple Constraints in Image Domain
摘要： X-ray spectral CT appears as a new promising imaging modality for the quantitative measurement of materials in an object, compared to conventional energy-integrating CT or dual energy CT. We consider material decomposition in spectral CT as an overcomplete ill-conditioned inverse problem. To solve the problem, we make full use of multi-dimensional nature and high correlation of multi-energy data and spatially neighboring pixels in spectral CT. Meanwhile, we also exploit the fact that material mass density has limited value. The material decomposition is then achieved by using bounded mass density, local joint sparsity and structural low-rank (DSR) in image domain. The results on numerical phantom demonstrate that the proposed DSR method leads to more accurate decomposition than usual pseudo-inverse method with singular value decomposition (SVD) and current popular sparse regularization method with (cid:2)1-norm constraint.
关键词： Sparse representation,X-ray spectral CT,Material decomposition,Low-rank representation
更新于2025-09-23 15:22:29
-
Robust Hyperspectral Image Domain Adaptation With Noisy Labels
摘要： In hyperspectral image (HSI) classification, domain adaptation (DA) methods have been proved effective to address unsatisfactory classification results caused by the distribution difference between training (i.e., source domain) and testing (i.e., target domain) pixels. However, these methods rely on accurate labels in source domain, and seldom consider the performance drop resulted by noisy label, which often happens since labeling pixel in HSI is a challenging task. To improve the robustness of DA method to label noise, we propose a new unsupervised HSI DA method, which is constructed from both feature-level and classifier-level. First, a linear transformation function is learned in feature-level to align the source (domain) subspace with the target (domain) subspace. Then, a robust low-rank representation based classifier is developed to well cope with the features obtained from the aligned subspace. Since both subspace alignment and the classifier are immune to noisy labels, the proposed method obtains good classification results when confronting with noisy labels in source domain. Experimental results on two DA benchmarks demonstrate the effectiveness of the proposed method.
关键词： low-rank representation,hyperspectral image (HSI) classification,Domain adaptation,subspace alignment
更新于2025-09-23 15:22:29
-
[IEEE 2019 North American Power Symposium (NAPS) - Wichita, KS, USA (2019.10.13-2019.10.15)] 2019 North American Power Symposium (NAPS) - Local Smart Inverter Control to Mitigate the Effects of Photovoltaic (PV) Generation Variability
摘要： Recently, position-patch-based face hallucination methods have received much attention, and obtained promising progresses due to their effectiveness and ef?ciency. A locality-constrained double low-rank representation (LCDLRR) method is proposed for effective face hallucination in this paper. LCDLRR attempts to directly use the image-matrix based regression model to compute the representation coef?cients to maintain the essential structural information. On the other hand, LCDLRR imposes a low-rank constraint on the representation coef?cients to adaptively select the training samples that belong to the same subspace as the inputs. Moreover, a locality constraint is also enforced to preserve the locality and the sparsity simultaneously. Compared with previous methods, our proposed LCDLRR considers locality manifold structure, cluster constraints, and structure error simultaneously. Extensive experimental results on standard face hallucination databases indicate that our proposed method outperforms some state-of-the-art algorithms in terms of both visual quantity and objective metrics.
关键词： low-rank representation,position-patch,Face hallucination,nuclear norm
更新于2025-09-23 15:21:01
-
[IEEE 2019 12th UK-Europe-China Workshop on Millimeter Waves and Terahertz Technologies (UCMMT) - London, United Kingdom (2019.8.20-2019.8.22)] 2019 12th UK-Europe-China Workshop on Millimeter Waves and Terahertz Technologies (UCMMT) - InGaAs/AlAs Metamorphic Asymmetric Spacer Tunnel (mASPAT) Diodes on GaAs Substrate for Microwave/millimetre-wave Applications
摘要： Recently, position-patch-based face hallucination methods have received much attention, and obtained promising progresses due to their effectiveness and efficiency. A locality-constrained double low-rank representation (LCDLRR) method is proposed for effective face hallucination in this paper. LCDLRR attempts to directly use the image-matrix based regression model to compute the representation coefficients to maintain the essential structural information. On the other hand, LCDLRR imposes a low-rank constraint on the representation coefficients to adaptively select the training samples that belong to the same subspace as the inputs. Moreover, a locality constraint is also enforced to preserve the locality and the sparsity simultaneously. Compared with previous methods, our proposed LCDLRR considers locality manifold structure, cluster constraints, and structure error simultaneously. Extensive experimental results on standard face hallucination databases indicate that our proposed method outperforms some state-of-the-art algorithms in terms of both visual quantity and objective metrics.
关键词： low-rank representation,position-patch,Face hallucination,nuclear norm
更新于2025-09-23 15:19:57
-
[IEEE 2019 IEEE 46th Photovoltaic Specialists Conference (PVSC) - Chicago, IL, USA (2019.6.16-2019.6.21)] 2019 IEEE 46th Photovoltaic Specialists Conference (PVSC) - Understanding and Mitigating the Contamination of Intrinsic poly-Si Gaps in Passivated IBC Solar Cells
摘要： Recently, position-patch-based face hallucination methods have received much attention, and obtained promising progresses due to their effectiveness and ef?ciency. A locality-constrained double low-rank representation (LCDLRR) method is proposed for effective face hallucination in this paper. LCDLRR attempts to directly use the image-matrix based regression model to compute the representation coef?cients to maintain the essential structural information. On the other hand, LCDLRR imposes a low-rank constraint on the representation coef?cients to adaptively select the training samples that belong to the same subspace as the inputs. Moreover, a locality constraint is also enforced to preserve the locality and the sparsity simultaneously. Compared with previous methods, our proposed LCDLRR considers locality manifold structure, cluster constraints, and structure error simultaneously. Extensive experimental results on standard face hallucination databases indicate that our proposed method outperforms some state-of-the-art algorithms in terms of both visual quantity and objective metrics.
关键词： position-patch,nuclear norm,Face hallucination,low-rank representation
更新于2025-09-23 15:19:57
-
[IEEE 2019 IEEE 16th International Conference on Group IV Photonics (GFP) - Singapore, Singapore (2019.8.28-2019.8.30)] 2019 IEEE 16th International Conference on Group IV Photonics (GFP) - Spectral Engineering of Photonic Filters Based on Mode Splitting in Self-Coupled Silicon Nanowire Waveguides
摘要： Recently, position-patch-based face hallucination methods have received much attention, and obtained promising progresses due to their effectiveness and ef?ciency. A locality-constrained double low-rank representation (LCDLRR) method is proposed for effective face hallucination in this paper. LCDLRR attempts to directly use the image-matrix based regression model to compute the representation coef?cients to maintain the essential structural information. On the other hand, LCDLRR imposes a low-rank constraint on the representation coef?cients to adaptively select the training samples that belong to the same subspace as the inputs. Moreover, a locality constraint is also enforced to preserve the locality and the sparsity simultaneously. Compared with previous methods, our proposed LCDLRR considers locality manifold structure, cluster constraints, and structure error simultaneously. Extensive experimental results on standard face hallucination databases indicate that our proposed method outperforms some state-of-the-art algorithms in terms of both visual quantity and objective metrics.
关键词： low-rank representation,position-patch,Face hallucination,nuclear norm
更新于2025-09-19 17:13:59
-
[ACM Press the 2019 International Conference - Wuhan, Hubei, China (2019.07.12-2019.07.13)] Proceedings of the 2019 International Conference on Artificial Intelligence and Computer Science - AICS 2019 - A Novel Infrared and Visible Image Fusion Using Low-rank Representation and Simplified Dual Channel Pulse Coupled Neural Network
摘要： This paper establishes a novel fusion scheme for infrared (IR) and visual (VI) images via low-rank representation (LRR), total variation (TV) model and simplified dual channel pulse coupled neural network (S-DPCNN) to effectively extract the major and salient information, which address some problems in existing fusion methods low-contrasting such as blurry edge, heterogeneous and information redundancy. The first step of the proposed method is to extract the valuable features of the IR images based on frequency tuned based- LRR (FT-LRR) algorithm aims to separate the corresponding salient regions and backgrounds. Furthermore, we adopt a choose-maximum rule to retain the significance information of the source images to the maximum extent for the salient region. For the background, IR and VI images are decomposed into a low-pass coefficient and a series of high-pass coefficients using non-subsampled shearlet transform (NSST). Then, the TV model is utilized to fuse the low-pass coefficient, in the meanwhile, the modified average gradient (MAG) is used to stimulate the S-DPCNN which aims to fuse high-pass coefficients. The fused background is obtained by taking inverse NSST. Finally, the robust fused image is generated by adding the fused salient region and background. Large amounts of experiment results and metrics demonstrate that the proposed framework exhibits good visual performance and has obvious superiorities over other state-of-the-art methods in both subjective and objective evaluation.
关键词： Image Fusion,Pulse Coupled Neural Network,Total Variation Model,Low-rank Representation
更新于2025-09-16 10:30:52
-
[IEEE 2018 International Conference On Advances in Communication and Computing Technology (ICACCT) - Sangamner, India (2018.2.8-2018.2.9)] 2018 International Conference On Advances in Communication and Computing Technology (ICACCT) - Low-Rank and Sparse Representation for Anomaly Detection in Hyperspectral Images
摘要： In view of low rank and scanty portrayal, a novel technique is projected for irregularity discovery in hyperspectral pictures. Working of this technique is rely upon partition of the base and the oddities information. Background word reference is utilized to around speak to every pixel out of sight and low rank network is utilized to speak to coefficients of all pixels, background part is show by utilizing low-rank initiating regularization term to portrayal coefficient I can better describe every pixel's nearby portrayal. Also, to make the lexicon more steady and discriminative, a lexicon development system is took into utilization. By the utilization of reaction of leftover grid peculiarities are resolved. Combination of both neighbourhood imperative and favourable position of my proposed calculation for hyperspectral pictures. I have hewed away at both mimicked and genuine informational indexes for leading tests and to get exploratory outcomes. My investigation finishes up being the finest and promising for irregularity discovery.
关键词： Hyperspectral image (HSI) examination,Robust principal component analysis(RPCA),low-rank representation (LRR) and sparse representation (SR),Anomaly recognition,dictionary building
更新于2025-09-10 09:29:36
-
Semi-Supervised hyperspectral image classification using local low-rank representation
摘要： In the area of hyperspectral image (HSI) classification, graph-based semi-supervised learning (SSL) has been proved to be highly effective. Constructing a proper graph is critical for graph-based SSL tasks. In HSI, spectral distance is widely used to calculate the weight of graph edge, though it can be influenced by noise and outliers. Meanwhile, links among all the data points are incorporated in the graph, including those from different subspaces. Thus the constructed graph might contain incorrect information. In this letter, a novel semi-supervised HSI classification method using local low-rank representation (SL2R) is proposed. Edge weight calculation will not be affected by noise or outliers thanks to the robustness of low-rank representation (LRR). Since each graph is constructed at local level, where pixels are basically embedded in the same subspace, links among uncorrelated pixels can be removed. Moreover, spatial context is naturally characterized by low-rank constraint on adjacent pixels. Experimental results on two data sets (Indian Pines and Botswana) confirm the effectiveness of the proposed method.
关键词： spectral-spatial classification,semi-supervised learning,hyperspectral image classification,low-rank representation
更新于2025-09-10 09:29:36
-
[IEEE IGARSS 2018 - 2018 IEEE International Geoscience and Remote Sensing Symposium - Valencia, Spain (2018.7.22-2018.7.27)] IGARSS 2018 - 2018 IEEE International Geoscience and Remote Sensing Symposium - Manifold Regularized Low-Rank Representation for Hyperspectral Anomaly Detection
摘要： A novel method for hyperspectral anomaly detection based on low-rank representation with manifold regularization is proposed in this paper. Usually, a hyperspectral imagery can be modeled as a superposition of two parts: background part with low rank dimensionality and anomaly part described by a sparse matrix. Low-rank representation (LRR) can be used to find the lowest rank representation of all pixels jointly which represents the background part, then the anomaly part is contained in the residual of the original image. To learn a more discriminative representation, we incorporate the manifold regularization term into the original LRR model. An important advantage of the proposed method is that it can utilize the global low rank property and local geometrical structure jointly. The experimental results on both simulated and real hyperspectral datasets validate the effectiveness of the proposed method.
关键词： local geometrical structure,manifold regularization,low-rank representation,anomaly detection,Hyperspectral imagery
更新于2025-09-10 09:29:36