Graph and Total Variation Regularized Low-Rank Representation for Hyperspectral Anomaly Detection
The DOI matches its Crossref record; 56 of the 56 references it cites were found in a registry or an index.
Abstract
Anomaly detection is of great importance among hyperspectral applications, which aims at locating targets that are spectrally different from their surrounding background. A variety of anomaly detection methods have been proposed in the past. However, most of them fail to take the high spectral correlations of all the pixels into consideration. Low-rank representation (LRR) has drawn a great deal of interest in recent years, as a promising model to exploit the intrinsic low-rank property of hyperspectral images. Nevertheless, the original LRR model only analyzes the spectral signatures without taking advantage of the valuable spatial information in hyperspectral images. Furthermore, it has been shown that the local geometrical information of the hyperspectral data is also important for discrimination between the anomalies and background pixels. In this article, we incorporate the graph regularization and total variation (TV) regularization into the LRR formulation and propose a novel anomaly detection method based on graph and TV regularized LRR (GTVLRR) model, to preserve the local geometrical structure and spatial relationships in hyperspectral images. Extensive experiments have been conducted on both simulated and real hyperspectral data sets. The experimental results demonstrate the superiority of the proposed method over conventional and state-of-the-art anomaly detection methods.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Citations
Cites
Boyd, S., Parikh, N., Chu, E., Peleato, B., & Eckstein, J. (2011) · Foundations and Trends® in Machine Learning · doi:10.1561/2200000016 · Cited by 17,075
Nonlinear total variation based noise removal algorithms
Rudin, L. I., Osher, S., & Fatemi, E. (1992) · Physica D: Nonlinear Phenomena · doi:10.1016/0167-2789(92)90242-f · Cited by 15,704
Nonlinear Dimensionality Reduction by Locally Linear Embedding
Roweis, S. T., & Saul, L. K. (2000) · Science · doi:10.1126/science.290.5500.2323 · Cited by 14,965
A Global Geometric Framework for Nonlinear Dimensionality Reduction
Tenenbaum, J. B., Silva, V. D., & Langford, J. C. (2000) · Science · doi:10.1126/science.290.5500.2319 · Cited by 13,585
Boyd, S. (2010) · now publishers, Inc. eBooks · doi:10.1561/9781601984616 · Cited by 12,677
Robust principal component analysis?
Candès, E. J., Li, X., Ma, Y., & Wright, J. (2011) · Journal of the ACM · doi:10.1145/1970392.1970395 · Cited by 6,921
A Singular Value Thresholding Algorithm for Matrix Completion
Cai, J.-F., Candès, E. J., & Shen, Z. (2010) · SIAM Journal on Optimization · doi:10.1137/080738970 · Cited by 5,809
Laplacian Eigenmaps and Spectral Techniques for Embedding and Clustering
Belkin, M., & Niyogi, P. (2002) · Advances in Neural Information Processing Systems 14 · doi:10.7551/mitpress/1120.003.0080 · Cited by 4,412
Robust Recovery of Subspace Structures by Low-Rank Representation
Liu, G., Lin, Z., Yan, S., Sun, J., Yu, Y., & Ma, Y. (2013) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2012.88 · Cited by 3,594
Kuybeda, O., Frank, G. A., Bartesaghi, A., Borgnia, M., Subramaniam, S., & Sapiro, G. (2013) · Journal of Structural Biology · doi:10.1016/j.jsb.2012.10.010 · Cited by 2,162
Research works printed but not linked: 1
- chung, 1997, Spectral Graph Theory
Cited by
Items 201–266 of 266
Hyperspectral Anomaly Detection via S 1/2 and Total Variation Low Rank Matrix Decomposition
Wang, J., Huang, P., Zhang, K., & Wang, Q. (2021) · IEEE Geoscience and Remote Sensing Letters · doi:10.1109/lgrs.2021.3096720 · Cited by 3
“The hyperparameters in GTVLRR are the same as this letter, i.e., λ = 0.05, β = 0.2, and γ = 0.02.”
BackgroundMethodInfluentialYang, J., Zhao, J., Chen, L., Ning, H., & Li, Y. (2025) · Remote Sensing · doi:10.3390/rs17183137 · Cited by 2
AdaptHAD: Adaptive One-Step Hybrid Network for Hyperspectral Anomaly Detection
Dai, X., Dong, Y., Zhang, Y., & Du, B. (2025) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2025.3625279 · Cited by 4
“Representation learning-based approaches aim to enhance adaptability to complex backgrounds and better capture non-linear data characteristics in some cases, such as collaborative representation (CR), sparse representation (SR), and LR decomposition [25], [26], [27].”
Küçük, F. (2026) · Scientific Reports · doi:10.1038/s41598-026-41223-w · Cited by 5
Zhao, Z., Song, J., Zhou, H., & Zhu, Y. (2026) · Pattern Recognition · doi:10.1016/j.patcog.2026.113925 · Cited by 3
Liu, Y., Lin, C.-H., & Kuo, Y.-C. (2022) · Workshop on Hyperspectral Image and Signal Processing: Evolution in Remote Sensing · doi:10.1109/whispers56178.2022.9955089 · Cited by 2
“Apart from RPCA model, another representative low-rank model applied in HAD is low-rank representation [4, 5], in which HSI pixels are approximated by the lowest rank representation in multiple subspaces.”
MethodLuyao, 李. L., Zhongwei, 李. L., Leiquan, 王. W., Juan, 李. L., & Shunxiao, 史. S. (2023) · Laser & Optoelectronics Progress · doi:10.3788/lop222705 · Cited by 2
Young, S.-S., Lin, C.-H., & Lin, J.-T. (2023) · Workshop on Hyperspectral Image and Signal Processing: Evolution in Remote Sensing · doi:10.1109/whispers61460.2023.10431118 · Cited by 2
“In this paper, the proposed CiDAR-Former is compared with other benchmark algorithms including LRX [4], GTVLRR [6], AED [17], PTA [7], and RGAE [9].”
BackgroundMethodZhao, Y., Su, H., Wu, Z., Xue, Z., & Du, Q. (2024) · IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing · doi:10.1109/jstars.2024.3476319 · Cited by 2
“Among these methods, statistical modeling-based methods and representation-based methods are the most representative ones [27], [28].”
MethodKaixing, 何. H., Zheng, 蒋. J., Bin, 刘. L., & Xiaokang, 张. Z. (2024) · Laser & Optoelectronics Progress · doi:10.3788/lop240872 · Cited by 2
Shahnas, P., & Malathy, S. (2025) · Chemometrics and Intelligent Laboratory Systems · doi:10.1016/j.chemolab.2025.105463 · Cited by 2
Ensemble and Random Collaborative Representation-Based Anomaly Detector for Hyperspectral ImageryPreprint
Wang, R., Lu, Y., Zhang, Q., Nie, F., Wang, Z., & Li, X. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2101.01976 · Cited by 2
“The value of ensemble size T is changed within the range of [5 , 10 , · · · , 50] and the number of random sampling r is set to 5 .”
Du, Z., Yang, L., & Tang, M. (2023) · Journal of Circuits Systems and Computers · doi:10.1142/s0218126624501457 · Cited by 3
Hyperspectral Anomaly Detection via Enhanced 3DTV and Sparse Reweighted Regularization
Xiao, Q., Zhao, L., Chen, S., & Li, X. (2024) · IEEE Geoscience and Remote Sensing Letters · doi:10.1109/lgrs.2024.3441612 · Cited by 1
“…algorithms are selected as comparison algorithms, including a statistics-based method LRX [3], a CR-based method CRD [6], a DL-based method RGAE [4], and nine LRaSD-based methods (LRASR [9], GTVLRR [10], LRSNCR [8], AHMID [16], LEBSR [11], PTA [14], PCA-TLRSR [13], LARTVAD [15], and GNLTR [12]).”
BackgroundMethodInfluentialInformation Entropy Estimation Based on Point-Set Topology for Hyperspectral Anomaly Detection
Sun, X., Zhuang, L., Gao, L., Gao, H., Sun, X., Liu, Y., Liu, Y., & Zhang, B. (2024) · IEEE International Geoscience and Remote Sensing Symposium (IGARSS) · doi:10.1109/igarss53475.2024.10641092 · Cited by 6
“…robust principal component analysis RXD (RPCA-RXD) [26]; collaborative representation-based detector (CRD) [11]; low-rank and sparse representation-based detector (LRASR) [12]; graph total variation regularized low rank representation-based detector (GTVLRR) [27]; and two-step GLRT (2S-GLRT) [28].”
MethodLi, Z., Wei, J., Wang, L., Yang, Y., & Wang, Y. (2025) · International Journal of Remote Sensing · doi:10.1080/01431161.2025.2459214 · Cited by 1
“As shown in Figure 4(d), GTVLRR, LREN, and GAED algorithms are significantly impacted by the presence of stripe noise bands, leading to messy and unreliable detection result maps.”
InfluentialSpectral Similarity on Low-Rank and Sparse Component for Anomaly Detection
Song, M., Guo, Z., & Bao, H. (2024) · Workshop on Hyperspectral Image and Signal Processing: Evolution in Remote Sensing · doi:10.1109/whispers65427.2024.10876453 · Cited by 2
Higher-Order Grouped Outlier Robust PCA for Spatio-temporal Anomaly Detection
Indibi, M., & Aviyente, S. (2025) · IEEE Statistical Signal Processing Workshop (SSP) · doi:10.1109/ssp64130.2025.11073198 · Cited by 1
“This is important for a variety of applications such as background-foreground separation in video [2], target detection in hyperspectral imaging (HSI) [3] or event detection from urban traffic data [4].”
ProTD: Prompted Tensor Decomposition for Hyperspectral Anomaly Detection
Fu, B.-Z., Huang, T., Zhao, X.-L., Wu, W.-H., & Peng, H. (2025) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2025.3613978 · Cited by 3
“These methods include two statistics-based methods (i.e., RX [8], KIFD [51]), three model-based methods (i.e., CRD [14], GTVLRR [52], PCA-TLRSR [19]), and three deep learning-based methods (i.e., Auto-AD [53], DFAN-HAD [40], GAED [37]).”
KDAD: Knowledge Distillation-Based Anomaly Detection for Thermal Infrared Hyperspectral Image
Zhao, E., Zhang, H., Qu, N., Wang, Y., & Zhao, Y. (2025) · IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing · doi:10.1109/jstars.2025.3622117 · Cited by 2
“Given that anomalous targets in hyperspectral data typically occupy only a limited number of pixels while exhibiting a sparse spatial distribution, anomalies can be detected utilizing low-rank matrix decomposition techniques [29], [30].”
Hyperspectral Anomaly Detection via Enhanced Low-Rank and Joint Saliency Prior
Xiao, Q., Huang, R., Chen, S., & Li, X. (2025) · IEEE Geoscience and Remote Sensing Letters · doi:10.1109/lgrs.2025.3626197 · Cited by 2
Kong, W., Shen, D., Ma, X., Du, Z., Gao, Q., & Wang, H. (2025) · IEEE International Geoscience and Remote Sensing Symposium (IGARSS) · doi:10.1109/igarss55030.2025.11243130 · Cited by 2
“Seven typical HAD methods are used to do the comparison: RX [2], LSMAD [16], TPCA [17], LRASR [18], GTVLRR [9], PTA [10]) and LREN [6].”
Zhao, X., Wang, Y., Yin, J., Li, W., 宗艳桃, & Vivone, G. (2026) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2026.3698646 · Cited by 2
“Low-rank representation methods [13], [14] decompose the entire HSI matrix into a low-rank background part and a sparse anomaly part.”
Yang, Y., Su, H., Wu, Z., Yang, H. J., & Du, Q. (2026) · International Journal of Applied Earth Observation and Geoinformation · doi:10.1016/j.jag.2026.105422 · Cited by 1
Multiscale Spectral Local Contrast Meets 3-D Residual Saliency for Hyperspectral Anomaly Detection
Hu, L., Yue, J., Zhao, D., & Yang, D. (2026) · IEEE Geoscience and Remote Sensing Letters · doi:10.1109/lgrs.2026.3718225 · Cited by 2
“3) Baseline Methods: 11 representative HAD methods are selected for comparison, including RX [3], PTA [4], GTVLRR [12], TLRSR [6], GNLTR [7], AHMID [8], LEBSR [9], Turbo-GoDec [11], GAE [5], Auto-AD [22], and GT-HAD [10].”
Reddy, M. S. K., & Rajput, D. S. (2021) · International Journal of Digital Crime and Forensics · doi:10.4018/ijdcf.20210901.oa3 · Cited by 1
“In 2020, Cheng and Wang (2020) have developed a novel hyperspectral anomaly detection method on the basis of the GTVLRR model, which was the hybridization of graph regularization and TV regularization.”
MethodWang, X., Wu, Z., Xu, Y., Wei, Z., & Xia, L. (2021) · Lecture notes in computer science · doi:10.1007/978-3-030-87358-5_20 · Cited by 1
Wu, Z., Li, L., Wang, B., & Hu, B. (2024) · IEEE International Geoscience and Remote Sensing Symposium (IGARSS) · doi:10.1109/igarss53475.2024.10640718 · Cited by 1
“One such HAD method detector is the GTVLRR which incorporates the graph and total variation (TV) [1] regularizations into the low-rank representation (LRR) on the background abundances.”
MethodSelf-citationMerging Total Variation Into Low-Rank Representation for Hyperspectral Anomaly Detection
Li, L., Wu, Z., Wang, B., & Hu, B. (2024) · IEEE International Geoscience and Remote Sensing Symposium (IGARSS) · doi:10.1109/igarss53475.2024.10640866 · Cited by 1
“Moreover, the graph and total variation regularized LRR (GTVLRR) [7] adds total variation (TV) regularity with graph regularity to the LRR to preserve the geometric structure and correlation between pixels in the data.”
BackgroundMethodSelf-citationMeng, L., Xu, Y., Shen, Q., & Chen, Y. (2025) · Signal Processing · doi:10.1016/j.sigpro.2024.109875 · Cited by 1
Hybrid Transformer-Cnn for Hyperspectral Anomaly DetectionPreprint
Zhao, M., Zheng, W., & Hu, J. (2025) · SSRN Electronic Journal · doi:10.2139/ssrn.5082115 · Cited by 1
Graph and Sparsity Regularized Decomposition Model for Hyperspectral Anomaly Detection
Wang, J., Liu, J., & Sun, S. (2021) · International Conference on Wireless Communications and Signal Processing (WCSP) · doi:10.1109/wcsp52459.2021.9613223
“The parameters of the GTVLRR and LRASR are set according to the recommended values [13], [21].”
MethodHyperspectral Anomaly Detection Using Tensor Low-Rank Representation
Wang, M., Hong, D., Wu, Y., Gao, L., & Huang, M. J. (2022) · Lecture notes in electrical engineering · doi:10.1007/978-981-19-6052-9_12 · Cited by 1
Wang, Z.-C., Liu, J., Shang, J., Dai, L.-Y., Zheng, C.-H., & Wang, J. (2022) · Lecture notes in computer science · doi:10.1007/978-3-031-23198-8_12
Wang, Z.-C., Liu, J., Shang, J., Dai, L.-Y., Zheng, C.-H., & Wang, J. (2023) · Journal of Computational Biology · doi:10.1089/cmb.2023.0077
“The previously mentioned methods are primarily concerned with the global structure of the sample data and therefore ignore the important role of local geometric information in subspace clustering (Cheng and Wang, 2020).”
Feng, M., & Shu, Q. (2025) · International Conference on Electrical, Electronic Information and Communication Engineering (EEICE) · doi:10.1109/eeice65049.2025.11033754
“Besides, some detectors based on LRSA have also been proposed for HAD in [14] and [15].”
Hyperspectral anomaly detection based on nonlocal group sparsity models with low-rank regularization
Li, M., Shen, D., Kong, W., Wang, H., & Ma, X. (2025) · Remote Sensing Letters · doi:10.1080/2150704x.2025.2551936
“Seven popular hyperspectral anomaly detection methods were selected as the comparison, and they are RX (Reed and Yu 1990), GTVLRR (Cheng and Wang 2020), TPCA (Chen, Yang, and Wang 2018), PCA-TLRSR (Wang et al. 2023), Super-RPCA (Lin and Lin 2024), LREN (Jiang et al. 2021) and DeCNN-AD (Fu X. Y. and…”
InfluentialCluster-infused low-rank subspace learning for robust multi-label classification
Zhu, Z., Zhou, C., Sun, S., Ntaye, E., Shen, X., & Liu, Z. (2025) · Applied Intelligence · doi:10.1007/s10489-025-06837-z
“In image processing [ 56–58 ], low-rank representations are used for denoising and compression, which are typically done by methods like singular value decomposition (SVD) and Principal Com - ponent Analysis (PCA).”
Enhanced transformer with integrating weight-masked CNNs for robust hyperspectral anomaly detection
Zou, C., Lin, Y., & Zou, C. (2026) · Signal Processing Image Communication · doi:10.1016/j.image.2026.117600
Sample Similarity-Guided Spatial-Spectral Masked Autoencoder for Hyperspectral Anomaly Detection
Zhao, Z., Zhang, J., Song, J., Xiang, P., Zhao, D., & Zhou, H. (2025) · IEEE International Geoscience and Remote Sensing Symposium (IGARSS) · doi:10.1109/igarss55030.2025.11242636
“(*Corresponding Author (hxzhou@mail.xidian.edu.cn)). methods [1–4], representation-based methods [5–12], and deep learning-based methods [13–21].”
Li, C., & Quan, Y. (2025) · IEEE International Geoscience and Remote Sensing Symposium (IGARSS) · doi:10.1109/igarss55030.2025.11243974
“They are RX[11], CRD[12], RPCA-RX[4], LSMAD[6], LRASR[7], GTVLRR[13] and LRSNCR[14], respectively.”
Hyperspectral Anomaly Detection Via Deep Prior Mamba Network
Li, L., Wan, C., Wang, B., & Hu, B. (2025) · IEEE International Geoscience and Remote Sensing Symposium (IGARSS) · doi:10.1109/igarss55030.2025.11243036
“The GTVLRR [7] introduces low-rank representation (LRR) and total variation (TV) regularization term to exploit the spatial smoothness and low-rankness of the background, respectively.”
Self-citationFlow-Based Density Estimation for Hyperspectral Anomaly Detection
Sheng, J., Li, X., & Chen, S. (2025) · IEEE International Geoscience and Remote Sensing Symposium (IGARSS) · doi:10.1109/igarss55030.2025.11242820
“To demonstrate the effectiveness of the proposed method, RXAD [7], GTVLRR [16], LSMAD [17], PTA [18], AutoAD [11], and RGAE [19] are selected as the comparison methods.”
Yang, Y., Xu, W., Shen, H., Zheng, H., Jiao, Y., Su, H., & Du, Q. (2025) · IEEE International Geoscience and Remote Sensing Symposium (IGARSS) · doi:10.1109/igarss55030.2025.11242949
“The proposed algorithm GCR-LRD is compared with the advanced methods, RX [4], CRD [6], LRASR [7], and the latest methods, graph and total variation regularized low-rank representation for HAD (GTVLRR) [13], and GAED [9].”
Accelerating Generative Adversarial Network for Hyperspectral Anomaly Detection
Yu, K., & Wu, Z. (2025) · IEEE International Geoscience and Remote Sensing Symposium (IGARSS) · doi:10.1109/igarss55030.2025.11243137
“The chosen algorithms include RX [1], LRASR [5], IEEPST [10], and GTVLRR [11] detectors.”
Flexible Background Modeling for Hyperspectral Anomaly Detection Using Convolutional Sparse Coding
Sato, K., Naganuma, K., & Ono, S. (2025) · IEEE International Geoscience and Remote Sensing Symposium (IGARSS) · doi:10.1109/igarss55030.2025.11242248
“…effectiveness of our method, we compare it with seven existing methods, including state-of-the-art ones: three statistics-based methods GRX [10], LRX [11], and 2S-GLRT [12], one deep learning-based method RGAE [13], and three decomposition-based methods GTVLRR [14], PCA-TLRSR [16], and HADHTV [17].”
Frequency Domain Mask Guided Diffusion Model for Hyperspectral Anomaly Detection
Zhang, G., Sun, T., Yin, J., Zhang, S., & Wu, Y. (2025) · IEEE International Geoscience and Remote Sensing Symposium (IGARSS) · doi:10.1109/igarss55030.2025.11243872 · Cited by 1
“B. Experiment Setup 1) Comparison Methods : Six state-of-the-art HAD techniques are selected for comparison with the proposed method, which includes the traditional methods RX [10], MsRFQFT [14] GTVLRR [30], and the DL-based methods Auto-AD [21], RGAE [31], and BSDM [32].”
Hyperspectral Target Detection Based on Multi-Scale Depthwise Separable Convolution
Wang, L., Geng, J., Xu, S., Liu, Q., Zhao, F., Xu, J., & Li, X. (2025) · International Conference on Electrical Engineering and Computer Technology (IC2ECT) · doi:10.1109/ic2ect66838.2025.11290940
“In this part, we evaluated the DSC-HTD method proposed in this paper on three commonly used hyperspectral datasets, San Diego, Abu-Urban-1 and Abu-Urban-2, and compared it with four most advanced methods, LRASR [8], GTVLRR [9], PTA [10] and HTD-IRN.”
Hyperspectral Anomaly Detection Based on Tensor Deep Subspace Prior Modeling
Shang, W., Meng, C., Peng, J., & Wu, Z. (2026) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2026.3661867
“To demonstrate the e ff ectiveness of our method, we compare the proposed ITSR-(cid:96) 1 with several representative approaches, including RX [52], RPCA [53], GTVLRR [23], E-3DTV [50], PTA [54], TRPCA [55], LRTDTV [27], as well as an ablation version referred to as ITSR.”
Wang, Y., Deng, C., Zhao, E., & Yu, C. (2026) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2026.3670890 · Cited by 3
“The traditional baselines include 2S-GLRT [56], GTVLRR [23], and AHMID [57], while the deep learning-based models include GAED [37], RGAE [35], Auto-AD [36], GIL-HAD [58], and NL2Net [59].”
InfluentialWang, Q., Wang, Y., Xiang, P., Prasad, S., & Zhao, D. (2026) · IEEE Signal Processing Letters · doi:10.1109/lsp.2026.3679284
“Representation-based methods include GTVLRR [9], CRD [11], and AHMID [10].”
QSME-Net: Quad-State Mamba-Enhanced Network for Hyperspectral Anomaly Detection
Chen, K., Tu, B., He, Y., Liu, B., Li, J., & Plaza, A. (2026) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2026.3681885 · Cited by 2
“We benchmark QSME-Net against nine methods categorized into statistics-based (RX [13], KIFD [50]), representation-based (CRD [34], GTVLRR [51], PTA [33], and PCA-TLRSR [52]), and deep-learning-based approaches (Auto-AD [10], NL2Net [53], and BockNet [54]).”
Sato, K., & Ono, S. (2026) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp55912.2026.11463992
D2G-Net: Disentangling Spatio-Spectral Priors for Cross-Domain Hyperspectral Anomaly Detection
Xu, J., Zhang, C., Yang, X., & Ma, X. (2025) · Workshop on Hyperspectral Image and Signal Processing: Evolution in Remote Sensing · doi:10.1109/whispers69515.2025.11501590 · Cited by 1
“While traditional methods, including the statistical Reed-Xiaoli (RX) detector [1] and various representation-based models such as CRD [2], LSMAD [3], PTA[4], GTVLRR[5], RPCA-RX[6], and LRASR [7], laid important groundwork, their performance is often constrained This work was supported in part by the Key Research and…”
Quantum Fuzzy Detection of Anomaly Objects Using Hyperspectral Generative Adversarial Learning
Lin, C.-H., Young, S.-S., & Langari, R. (2025) · Workshop on Hyperspectral Image and Signal Processing: Evolution in Remote Sensing · doi:10.1109/whispers69515.2025.11501601
“…making HAD a fuzzy problem; 2) the mainstream optimization-based detectors model complex real-world scenarios using solely handcrafted properties [7], thereby limiting their practicality; 3) even with the powerful deep learning (DL) techniques, they only focus on background reconstruction under…”
Li, D., Lv, Y., Fanqiang, K., & Wang, Q. (2026) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2026.3697018 · Cited by 1
“In the HAD task, we select six comparison methods, including four LRASR methods (LSMAD [19], LRASR [42], PTA [43], and GTVLRR [47]), two tensor-based methods (TPCA [44] and PCA-TLRSR [46]), and one deep-learning–based method, DFAN [21].”
Xu, M., Wang, Q., Zheng, H., Kong, R., Liu, S., Sheng, H., & Li, Z. (2026) · IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing · doi:10.1109/jstars.2026.3710479
“Methods, such as LSMAD and LRASR, rely on the low-rank nature of HSI, separating the data into a low-rank background component and sparse anomalies [22], [23], [24], [25], [26].”
Kernel Bayesian Robust Tensor Factorization for Hyperspectral Anomaly Detection
Feng, Q., Kong, W., Zhao, X.-L., & Wang, J. (2026) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2026.3713792 · Cited by 1
“The comparison includes one classic statistical detector (RX [25]), three matrix-based methods (LSMAD [40], LRASR [42], and GTVLRR [43]), three tensor-based methods (PTA [55], PCA-TLRSR [52], and CF2-GNBRL [18]), one deep learning-based method (GAED [37]), and one Bayesian method (LSDM-MoG [41]).”
Jiang, X., Qiao, L., Gong, M., Li, J., Zhou, Y., & Wang, S. (2026) · International Journal of Remote Sensing · doi:10.1080/01431161.2026.2707261
“Moreover, as standard LRR primarily exploits spectral features, Cheng and Wang (2019) devised a graph and total variation-regularized LRR (GTVLRR) to effectively incorporate spatial information.”
Hu, L., Yue, J., Zhao, D., & Yang, D. (2026) · SSRN Electronic Journal · doi:10.2139/ssrn.7344817
Spatial-spectral attention-enhanced U-shaped blind spot network for hyperspectral anomaly detection
Zhao, Z., Song, J., Zhou, H., & Yang, T. (2026) · Neural Networks · doi:10.1016/j.neunet.2026.109548
Dual Prompt Guidance for Cross-Scene Hyperspectral Anomaly Detection
Li, G., Liu, L., Wang, Y., Xiao, C., Cuicui, H., & Li, K. (2026) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2026.3729813
“In addition, Cheng and Wang [43] introduce graph regularization to characterize the structural features of HSIs, enhancing the separability between background and anomalies.”
Xu, J., Zhang, C., Wang, X., Yang, X., Han, Y., & Ma, X. (2026) · Neural Networks · doi:10.1016/j.neunet.2026.109674
SRBNet: Selective Reconstruction Bypass for Edge-Robust Hyperspectral Anomaly Detection
Zhang, G., Sun, T., Zhang, S., He, J., & Xiong, Z. (2026) · IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing · doi:10.1109/jstars.2026.3733743
“To address this challenge, classical HAD has progressed from statistical detectors, such as Reed–Xiaoli (RX) [8], to representation-based and graph-regularized formulations [9], [10] and [11].”
Xu, J., Sun, L., Yuan, Z., Wang, M., & Xie, F. (2026) · IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing · doi:10.1109/jstars.2026.3739337
“These methods include five non-DNN-based methods including RX [7], LRX [8], GTVLRR [25], MsRFQFT [27], and PCA-TLRSR[26]andfourDNN-basedmodelsincludingGAED[44], RGAE [45], Auto-AD [46], GT-HAD [47], FCMMA [50], and FS2CCTrans [51].”
A Tail-Energy Perturbed Black-Box Adversarial Attack for Anti-Detection of Hyperspectral Anomaly
Young, S.-S., Lin, C.-H., & Hsu, S. (2026) · IEEE International Geoscience and Remote Sensing Symposium (IGARSS) · doi:10.1109/igarss60577.2026.11731197
“Additionally, unlike the HC approaches that are predominantly DL-based classifiers and therefore easily affected by the model vulnerability, many effective HAD detectors are non-DL-based, including statistical-based [17], frequency-based [18], and low-rank representation (LRR)-based [19] algorithms.”
Metadata
- Authors
- Cheng, T., & Wang, B.
- Affiliations
- Fudan University (CN)
- Venue
- IEEE Transactions on Geoscience and Remote Sensing
- Published
- 2020-01
- Volume
- 58
- Issue
- 1
- Pages
- 391-406
- ISSN
- 0196-2892 (ISSN-L)
- 1558-0644
- OpenAlex
- W2972480129
- Cited by
- 266
- References
- 56
- Citations by year
- 2026: 32 · 2025: 68 · 2024: 54 · 2023: 42 · 2022: 41 · 2021: 25 · 2020: 4
- FWCI
- 11.73
- Type
- journal article
- Language
- en
- Open access
- closed
- Licence
- Locations
- IEEE Transactions on Geoscience and Remote Sensing (publishedVersion)
- Funders
- National Natural Science Foundation of China (61971141, 61572133, 61731021)
- Topics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Remote-Sensing Image Classification
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Infrared Target Detection Methodologies
- Physical Sciences › Computer Science › Artificial Intelligence › Anomaly Detection Techniques and Applications
- Keywords
- hyperspectral anomaly detection, low-rank representation, graph regularization, total variation regularization, geospatial data, spectral signatures
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.