Pith. sign in

REVIEW 5 major objections 5 minor 44 references

Adaptive Homophily Clustering: Structure Homophily Graph Learning with Adaptive Filter for Hyperspectral Image

T0 review · 5 major / 5 minor · reviewed 2026-08-10 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read This paper claims that adaptive filters plus pseudo-label-driven edge rewiring make the graph itself a trainable object, yielding the best reported clustering accuracy on three hyperspectral benchmarks.

desk verdict Strong OA numbers on three HSI benchmarks, but the paper's own tables contradict its 'best on all metrics' claim, and the graph-update loop is self-referential. read the letter →

arxiv 2501.01595 v2 pith:4Z22XZP4 submitted 2025-01-03 cs.CV

Yao Ding , Weijie Kang , Aitao Yang , Zhili Zhang , Junyang Zhao , Jie Feng , Danfeng Hong , Qinhe Zheng This is my paper · ORCID
classification cs.CV
keywords hyperspectralimageclusteringadaptivegraphfilterhomophily-enhancedstructurelearningself-trainingedgesparsificationsuperpixeljointnetworkoptimizationunsupervisedland-covermapping
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper tries to establish that hyperspectral image clustering with zero labeled pixels improves substantially when the graph is not fixed but is rewritten as the clustering proceeds. The proposed AHSGC method (a homophily structure graph-learning framework with an adaptive filter) segments the image into superpixels, encodes them with an adaptive graph filter that learns to keep both low-frequency similarity signals and high-frequency difference signals, and then sparsifies graph edges from its own confident pseudo-labels: edges are added inside clusters and removed between clusters. It reports the best clustering accuracy among the methods compared on all three benchmarks, with overall accuracies of 83.60% on Salinas, 63.65% on Pavia University, and 86.03% on Trento, while keeping training time lower than several deep clustering baselines. If this holds, land-cover maps for new hyperspectral scenes could be produced without any labeled training samples and with modest compute.

What carries the argument

The central object carrying the argument is the homophily-enhanced structure-learning loop coupled with an adaptive graph filter. The adaptive filter, $F_A=\mu(S_{rw})^k X+(1-\mu)(I-S_{rw})^k X$, replaces the standard low-pass-only graph convolution with a learnable mixture of random-walk-normalized propagation and its Laplacian complement, so the encoder can preserve both smooth and non-smooth node features. The structure-learning loop uses pseudo-labels and high-confidence node subsets to estimate node similarities in embedding space, then rewrites the graph by intra-cluster edge recovery and inter-cluster edge removal. The rewritten adjacency $\bar{A}$ enters the reconstruction loss, which is what trains the embeddings toward the task-specific graph; the whole joint objective $\mathcal{L}_O$ is what makes the graph update and the network training happen in one loop.

What would settle it

Run AHSGC on Salinas with the same number of added and removed edges, but choose the added and removed edge sets at random instead of by confidence-ranked pseudo-labels; if overall accuracy stays near the reported 83.60%, the pseudo-label guidance is not the source of the gain.

Watch

Extended reading notes

Core claim

For an unlabeled hyperspectral image, AHSGC constructs a superpixel-level graph and learns node embeddings through a stack of adaptive graph filters, $F_A=\mu(S_{rw})^k X+(1-\mu)(I-S_{rw})^k X$, so the encoder can keep both similarity-bearing low-frequency content and difference-bearing high-frequency content. A self-training decoder turns embeddings into soft cluster assignments $q_{ik}$ via a Student's $t$-distribution and sharpens them with KL divergence, giving pseudo-labels $c_i=\arg\max_k q_{ik}$. The homophily-enhanced structure-learning module then takes the top-$\gamma$ most confident nodes per cluster, estimates pairwise similarities $S_{ij}^k=Z_i^k (Z_j^k)^T$ inside those subsets, adds edges among the most similar same-label pairs, removes edges between high-similarity different-label pairs, and updates the adjacency to $\bar{A}=A-A_{\mathcal{E}_{rm}}+A_{\mathcal{E}_{rc}}$. A reconstruction loss $\mathcal{L}_g=\|ZZ^T-\bar{A}\|_F^2$ trains the embeddings to reproduce this sparsified structure, and the combined objective $\mathcal{L}_O=\mathcal{L}_c+\mathcal{L}_g$ is optimized jointly before K-means on the final embeddings. The central claim is that this feedback loop corrects erroneous graph connections and extracts features that fixed-graph deep clustering methods miss.

Load-bearing premise

The load-bearing premise is that the pseudo-labels produced by the current clustering are accurate enough for the confidence-filtered edge additions and removals to improve the graph, because those same pseudo-labels are the only guide for rewriting the adjacency and there is no external correction in the loop.

Editorial extensions

If this is right

  • If the reported numbers hold, new hyperspectral scenes can be clustered into land-cover classes with no labeled pixels, with overall accuracy above 83% on Salinas and 86% on Trento.
  • The paper's ablations show each module contributes: removing the homogeneous-region generation, the adaptive filter encoder, or the homophily-enhanced structure learning lowers the five metrics on all three datasets.
  • Because the final step is K-means on embeddings trained to match the sparsified graph, the inference step is cheap; the reported test times are under a tenth of a second on all three datasets.
  • The adaptive-filter ablation supports the paper's claim that high-frequency boundary and difference information is a useful signal for HSI clustering, not just noise to be filtered out.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Because the graph update is driven by the clustering's own pseudo-labels, early confident mistakes are likely to be amplified rather than corrected; injecting controlled label noise in the first iterations and measuring recovery would test this.
  • The adaptive-filter mixing idea is not tied to hyperspectral images and could be carried to general graph clustering or semi-supervised node classification, where the low- versus high-frequency balance is usually fixed by architecture rather than learned.
  • The edge-sparsification rules assume homophily, so on heterophilic scenes or graphs they may remove useful cross-class edges; the reported gains may not transfer without retuning the recovery and removal ratios.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

5 major / 5 minor

Summary. The manuscript proposes AHSGC, an unsupervised clustering method for hyperspectral images that combines superpixel-based graph construction, an adaptive-filter graph encoder for low- and high-frequency feature extraction, a self-training clustering decoder based on KL divergence, and a homophily-enhanced structure-learning module that sparsifies and rewires the graph during training. The method is evaluated on Salinas, Pavia University, and Trento, compared against nine clustering baselines, and reported to achieve higher OA, Kappa, NMI, ARI, and Purity on most settings, along with lower computational cost. Ablation studies and hyperparameter sensitivity analyses are also provided.

Significance. If the empirical claims were fully supported, AHSGC would be a practically useful contribution to unsupervised HSI clustering: it combines several plausible ingredients (superpixel graphs, adaptive filtering, dynamic graph sparsification, joint self-training) and the reported runtime/complexity advantages are attractive. The paper also ships an extensive comparison, ablation experiments, and a promise of code release. However, the central 'best clustering performance' claim is internally contradicted by the paper's own tables on two of the three datasets, and the self-training graph-update loop has a circularity that is not addressed. These issues are load-bearing because the headline contribution is precisely the claimed state-of-the-art accuracy and the adaptive graph-updating mechanism.

major comments (5)
  1. [Section IV-C.1, Table IV] The text states that AHSGC 'achieves the best clustering performance with 83.60%, 81.62%, 77.42%, and 83.68% in terms of OA, Kappa, NMI, ARI, and Purity' on Salinas and claims a 0.73% improvement on NMI. This is contradicted by Table IV: AHSGC has NMI = 0.8587, while EGAE has NMI = 0.8769, so AHSGC is not the best on NMI and the alleged gain is actually a 1.82-percentage-point deficit. The stated 0.73% improvement appears to use NCSC (0.8514) as the second-best method rather than the actual second-best EGAE. The narrative must be corrected to compare against EGAE and to restrict the 'best' claim to the metrics that actually support it.
  2. [Section IV-C.2, Table V] The claim 'our AHSGC method still performs the best among all investigated clustering methods' on Pavia University is not supported by Table V. AHSGC has NMI = 0.5251, which is lower than both K-means (0.5529) and EGAE (0.5372), and Purity = 0.6980, which is lower than EGAE (0.7126). The conclusion that AHSGC is best across the board is therefore false as written; the empirical summary must identify the specific metrics and datasets for which AHSGC is actually leading.
  3. [Section III-D, Eqs. (24)-(25)] The graph reconstruction loss L_g = ||ZZ^T - A̅||_F^2 trains Z to match A̅, but A̅ is constructed in Eq. (24) from edge sets obtained via Eqs. (20)-(23) using S = ZZ^T and pseudo-labels c_i = argmax_k q_ik computed from the same Z. The reconstruction target is therefore a thresholded function of the model's own output, not an independent supervisory graph. If the high-confidence pseudo-labels are systematically wrong, the update loop can reinforce errors instead of correcting them. This circularity is load-bearing for the homophily-enhanced structure-learning contribution; the authors should either provide an independent reconstruction target or report a controlled experiment (e.g., randomizing or corrupting the pseudo-labels) showing that the loop does not merely chase its own output.
  4. [Section IV-B.3 and Section IV-F] The paper states that all methods are run ten times to eliminate bias, but Tables IV-VI report only point estimates with no standard deviations, confidence intervals, or significance tests. Moreover, the hyperparameters (xi, eta, T, L) are selected by grid search on the test set, as described in Section IV-F. Several reported advantages are very small (e.g., Salinas Purity 0.8368 for AHSGC versus 0.8357 for EGAE), so without variance estimates or a validation-based selection protocol these differences are not established. Please report mean and standard deviation over the ten runs and separate validation data from the test set for hyperparameter selection.
  5. [Section III-B, Eq. (11)] The adaptive filter in Eq. (11) depends on a parameter mu that is described as a learnable balance between high-pass and low-pass information, but no update rule, initialization, or constraint for mu is provided. Table III lists seven preset parameters yet omits mu, and the parameter analysis in Section IV-F does not discuss it. Since the 'adaptive filter' is a central contribution, the training or selection of mu must be specified.
minor comments (5)
  1. [Section IV heading] The section title 'EXPLEMENTS' should be corrected to 'EXPERIMENTS'.
  2. [Figures in Sections IV-F through IV-H] Three distinct figures are all numbered 'Fig. 8': the parameter-sensitivity plots, the t-SNE visualization, and the ablation bar charts. These should be renumbered sequentially.
  3. [Table I and Eq. (13)] The notation for Z is inconsistent: Table I defines Z as the pixel-superpixel correlation matrix in R^{hw x N}, while Eq. (13) uses Z for the graph encoder output in R^{N x D}. Please use distinct symbols or clarify the reuse.
  4. [Section III-D, Eq. (23)] The text refers to eta as the 'inter-cluster edge removal ratio', but Eq. (23) describes it as the percentage of retained edges, and Table I calls it the removal ratio. This ambiguity should be resolved, as the interpretation affects the graph sparsification behavior.
  5. [Section IV-C.1] The sentence reporting Salinas results lists four numerical values for five metrics; please report all five values explicitly and consistently with Table IV.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the self-referential graph-reconstruction loss is a self-training regularizer, and the paper's central accuracy claim is empirical rather than derived.

full rationale

The method is a standard self-training clustering pipeline, not a first-principles derivation whose output is secretly its input. The graph reconstruction loss L_g = ||ZZ^T - A̅||_F^2 (Eq 25) does use a target A̅ that is rebuilt each iteration from the same model output: A̅ = A - A_{E_rm} + A_{E_rc} (Eq 24), with E_rc selected by ranking S_ij = Z_i Z_j^T (Eqs 19-21) and E_rm by pseudo-labels c_i = argmax_k q_ik (Eqs 17, 23). This is indeed a self-referential, potentially error-reinforcing update loop if high-confidence pseudo-labels are systematically wrong, and it is a legitimate robustness limitation. But it is not circularity under the analysis rules: no external quantity is being predicted from a fit of itself; the pseudo-graph is explicitly a training target ('p_ik is treated as ground-truth', Section III-E), and the paper's headline contribution is an empirical clustering-accuracy claim measured against ground truth (Tables IV-VI), not a quantity forced by construction. The citations to the authors' earlier graph-clustering works are contextual, and no load-bearing uniqueness or first-principles theorem is imported from them. The Section IV-C.1 claim that AHSGC is best on all five metrics is contradicted by the paper's own Tables IV-V on NMI (EGAE 0.8769 vs AHSGC 0.8587 on Salinas) and on NMI/Purity for PU (k-means 0.5529 and EGAE 0.7126), but that is a reporting/correctness inconsistency, not circularity. Therefore no circular step is established.

Assumptions & free parameters 10 free parameters · 5 assumptions · 0 invented entities

The central claim rests on a large set of tuned hyperparameters (gamma, xi, eta, T, L, N, rho) and on the self-referential graph reconstruction target A̅. No independent entity is introduced, but the graph update loop lacks an external benchmark: the target graph is derived from the model's own embeddings and pseudo-labels.

free parameters (10)
  • mu (adaptive filter balance) = learned during training (value not reported)
    Eq (11): F_A = mu*(S_rw)^k X + (1-mu)*(I-S_rw)^k X; learned by gradient descent to balance low-pass and high-pass information.
  • gamma (high-confidence node extraction ratio) = 0.3 (Salinas, Trento), 0.5 (PU)
    Eq (18): selects top-gamma nodes per cluster for graph update; tuned per dataset in Table III.
  • xi (intra-cluster edge recovery ratio) = 0.5 (all datasets)
    Eq (21): adds edges among top-xi within-cluster node pairs; selected per dataset in Table III.
  • eta (inter-cluster edge removal ratio) = 0.05 (all datasets)
    Eq (23): removes low-ranked inter-cluster edges; Table III. The text also calls eta the retained-edge percentage, an internal inconsistency.
  • T (training iterations) = 50 (all datasets)
    Algorithm 1 and Table III; chosen by parameter analysis in Section IV-F.
  • L (learning rate) = 5e-4 (all datasets)
    Table III; chosen by parameter analysis in Section IV-F.
  • N (number of superpixels) = 580 (Salinas), 800 (PU), 550 (Trento)
    Table III; controls graph size and granularity; chosen per dataset without a stated criterion.
  • l or t (encoder layers) = 5
    Table III lists encoder layer number l; Algorithm 1 uses t layers; no sensitivity analysis for this parameter is shown.
  • rho (adjacency kernel bandwidth) = 0.2
    Eq (28): A_ij = exp(-rho*||X_i-X_j||^2) for neighbors; fixed by hand without ablation.
  • K (number of clusters) = 16 (SA), 9 (PU), 6 (Trento)
    Taken from dataset ground-truth class counts; clustering methods require K, but using label-derived K is a mild form of supervision.
assumptions (5)
  • standard math Graph Fourier convolution formulas (Eq 1-4) apply to the constructed superpixel graph.
    Used to justify the adaptive filter in Eq (11); this is standard spectral graph theory.
  • domain assumption SLIC superpixels are class-homogeneous, so averaging pixel spectra into node features preserves discriminative information.
    Section III-F: node features are mean spectra of pixels in each superpixel; if superpixels straddle land-cover boundaries, node features are contaminated.
  • domain assumption The HSI graph is homophilic: same-class nodes are more likely to be connected.
    Eq (5) defines homophily; Section III-D uses pseudo-labels to add/remove edges based on this assumption.
  • domain assumption Pseudo-labels from Student-t soft assignments are sufficiently reliable to guide edge updates.
    Eq (17)-(18) select confident nodes; Eqs (21)-(24) update the graph from these pseudo-labels.
  • ad hoc to paper The reconstructed adjacency A̅ computed from current embeddings Z is a valid supervisory target for L_g.
    Eq (24)-(25): L_g = ||ZZ^T - A̅||_F^2 with A̅ a thresholded function of ZZ^T; no external reference or justification prevents degenerate fixed points.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Adaptive Homophily Clustering: Structure Homophily Graph Learning with Adaptive Filter for Hyperspectral Image." pith.science (2026). https://pith.science/paper/4Z22XZP4

@misc{pith2026250101595,
  author       = {Pith},
  title        = {Pith review of: Adaptive Homophily Clustering: Structure Homophily Graph Learning with Adaptive Filter for Hyperspectral Image},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/4Z22XZP4}},
  note         = {Machine review of arXiv:2501.01595}
}
read the original abstract

Hyperspectral image (HSI) clustering has been a fundamental but challenging task with zero training labels. Currently, some deep graph clustering methods have been successfully explored for HSI due to their outstanding performance in effective spatial structural information encoding. Nevertheless, insufficient structural information utilization, poor feature presentation ability, and weak graph update capability limit their performance. Thus, in this paper, a homophily structure graph learning with an adaptive filter clustering method (AHSGC) for HSI is proposed. Specifically, homogeneous region generation is first developed for HSI processing and constructing the original graph. Afterward, an adaptive filter graph encoder is designed to adaptively capture the high and low frequency features on the graph for subsequence processing. Then, a graph embedding clustering self-training decoder is developed with KL Divergence, with which the pseudo-label is generated for network training. Meanwhile, homophily-enhanced structure learning is introduced to update the graph according to the clustering task, in which the orient correlation estimation is adopted to estimate the node connection, and graph edge sparsification is designed to adjust the edges in the graph dynamically. Finally, a joint network optimization is introduced to achieve network self-training and update the graph. The K-means is adopted to express the latent features. Extensive experiments and repeated comparative analysis have verified that our AHSGC contains high clustering accuracy, low computational complexity, and strong robustness. The code source will be available at https://github.com/DY-HYX.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

44 extracted references · 40 canonical work pages

  1. [1]

    QTN: Quaternion transformer network for hyperspectral image classification,

    X. Yang, W. Cao, Y. Lu, and Y. Zhou, “QTN: Quaternion transformer network for hyperspectral image classification,” IEEE Transactions on Circuits and Systems for Video Technology, vol. 33, no. 12, pp. 7370-7384, 2023

    work page 2023

  2. [2]

    SpectralGPT: Spectral remote sensing foundation model,

    D. Hong, B. Zhang, X. Li, Y. Li, C. Li, J. Yao, N. Yokoya, H. Li, P. Ghamisi, and X. Jia, “SpectralGPT: Spectral remote sensing foundation model,” IEEE Transactions on Pattern Analysis and Machine Intelligence, 2024

    work page 2024

  3. [3]

    Interactive Spectral -Spatial Transformer for Hyperspectral Image Classification,

    L. Song, Z. Feng, S. Yang, X. Zhang, and L. Jiao, “Interactive Spectral -Spatial Transformer for Hyperspectral Image Classification,” IEEE Transactions on Circuits and Systems for Video Technology, 2024

    work page 2024

  4. [4]

    Multiscale densely-connected fusion networks for hyperspectral images classification,

    J. Xie, N. He, L. Fang, and P. Ghamisi, “Multiscale densely-connected fusion networks for hyperspectral images classification,” IEEE Transactions on Circuits and Systems for Video Technology, vol. 31, no. 1, pp. 246-259, 2020

    work page 2020

  5. [5]

    MR -selection: A meta -reinforcement learning approach for zero -shot hyperspectral band selection,

    J. Feng, G. Bai, D. Li, X. Zhang, R. Shang, and L. Jiao, “MR -selection: A meta -reinforcement learning approach for zero -shot hyperspectral band selection,” IEEE Transactions on Geoscience and Remote Sensing, vol. 61, pp. 1-20, 2022

    work page 2022

  6. [6]

    Deep hybrid: multi - graph neural network collaboration for hyperspectral image classification,

    D. Yao, Z. Zhi -li, Z. Xiao -feng, C. Wei, H. Fang, C. Yao-ming, and W. -W. Cai, “Deep hybrid: multi - graph neural network collaboration for hyperspectral image classification,” Defence Technology, vol. 23, pp. 164-176, 2023

    work page 2023

  7. [7]

    Land use/land cover (LULC) classification using hyperspectral images: a review,

    C. Lou, M. A. Al -qaness, D. AL-Alimi, A. Dahou, M. Abd Elaziz, L. Abualigah, and A. A. Ewees, “Land use/land cover (LULC) classification using hyperspectral images: a review,” Geo-spatial Information Science, pp. 1-42, 2024

    work page 2024

  8. [8]

    Learning Unified Anchor Graph for Joint Clustering of Hyperspectral and LiDAR Data,

    Y. Cai, Z. Zhang, X. Liu, Y. Ding, F. Li, and J. Tan, “Learning Unified Anchor Graph for Joint Clustering of Hyperspectral and LiDAR Data,” IEEE Transactions on Neural Networks and Learning Systems, 2024

    work page 2024

Show all 44 references
  1. [9]

    Graph sample and aggregate -attention network for hyperspectral image classification,

    Y. Ding, X. Zhao, Z. Zhang, W. Cai, and N. Yang, “Graph sample and aggregate -attention network for hyperspectral image classification,” IEEE Geoscience and Remote Sensing Letters, vol. 19, pp. 1-5, 2021

    2021

  2. [10]

    Unsupervised self - correlated learning smoothy enhanced locality preserving graph convolution embedding clustering for hyperspectral images,

    Y. Ding, Z. Zhang, X. Zhao, W. Cai, N. Yang, H. Hu, X. Huang, Y. Cao, and W. Cai, “Unsupervised self - correlated learning smoothy enhanced locality preserving graph convolution embedding clustering for hyperspectral images,” IEEE Transactions on Geoscience Remote Sensing, vol...

    2022

  3. [11]

    IDA: Improving distribution analysis for reducing data complexity and dimensionality in hyperspectral images,

    A.-A. Dalal, M. A. Al -qaness, Z. Cai, and E. A. Alawamy, “IDA: Improving distribution analysis for reducing data complexity and dimensionality in hyperspectral images,” Pattern Recognition, vol. 134, pp. 109096, 2023

    2023

  4. [12]

    An efficient k - means clustering algorithm: Analysis and implementation,

    T. Kanungo, D. M. Mount, N. S. Netanyahu, C. D. Piatko, R. Silverman, and A. Y. Wu, “An efficient k - means clustering algorithm: Analysis and implementation,” IEEE transactions on pattern analysis and machine intelligence, vol. 24, no. 7, pp. 881-892, 2002

    2002

  5. [13]

    Fuzzy c-multiple-means clustering for hyperspectral image,

    X. Yang, M. Zhu, B. Sun, Z. Wang, and F. Nie, “Fuzzy c-multiple-means clustering for hyperspectral image,” IEEE Geoscience Remote Sensing Letters, vol. 20, pp. 1-5, 2023

    2023

  6. [14]

    On spectral clustering: Analysis and an algorithm,

    A. Ng, M. Jordan, and Y. Weiss, “On spectral clustering: Analysis and an algorithm,” Advances in neural information processing systems, vol. 14, 2001

    2001

  7. [15]

    The possibilistic c-means algorithm: insights and recommendations,

    R. Krishnapuram, and J. M. Keller, “The possibilistic c-means algorithm: insights and recommendations,” IEEE transactions on Fuzzy Systems, vol. 4, no. 3, pp. 385-393, 1996

    1996

  8. [16]

    Clustering by fast search and find of density peaks,

    A. Rodriguez, and A. Laio, “Clustering by fast search and find of density peaks,” science, vol. 344, no. 6191, pp. 1492-1496, 2014

    2014

  9. [17]

    Superpixel contracted neighborhood contrastive subspace clustering network for hyperspectral images,

    Y. Cai, Z. Zhang, P. Ghamisi, Y. Ding, X. Liu, Z. Cai, and R. Gloaguen, “Superpixel contracted neighborhood contrastive subspace clustering network for hyperspectral images,” IEEE Transactions on Geoscience and Remote Sensing, vol. 60, pp. 1-13, 2022

    2022

  10. [18]

    A unified framework for representation -based subspace clustering of out -of-sample and large -scale data,

    X. Peng, H. Tang, L. Zhang, Z. Yi, and S. Xiao, “A unified framework for representation -based subspace clustering of out -of-sample and large -scale data,” IEEE transactions on neural networks and learning systems, vol. 27, no. 12, pp. 2499-2512, 2015

    2015

  11. [19]

    Hierarchical clustering of hyperspectral images using rank -two nonnegative matrix factorization,

    N. Gillis, D. Kuang, and H. Park, “Hierarchical clustering of hyperspectral images using rank -two nonnegative matrix factorization,” IEEE Transactions on Geoscience and Remote Sensing, vol. 53, no. 4, pp. 2066-2078, 2014

    2014

  12. [20]

    Low -rank and sparse representation for hyperspectral image processing: A review,

    J. Peng, W. Sun, H. -C. Li, W. Li, X. Meng, C. Ge, and Q. Du, “Low -rank and sparse representation for hyperspectral image processing: A review,” IEEE Geoscience and Remote Sensing Magazine, vol. 10, no. 1, pp. 10-43, 2021

    2021

  13. [21]

    NSCKL: Normalized spectral clustering with kernel -based learning for semisupervised hyperspectral image classification,

    Y. Su, L. Gao, M. Jiang, A. Plaza, X. Sun, and B. Zhang, “NSCKL: Normalized spectral clustering with kernel -based learning for semisupervised hyperspectral image classification,” IEEE Transactions on Cybernetics, vol. 53, no. 10, pp. 6649-6662, 2022

    2022

  14. [22]

    Enhanced spatially constrained remotely sensed imagery classification using a fuzzy local double neighborhood information c -means clustering algorithm,

    H. Zhang, L. Bruzzone, W. Shi, M. Hao, and Y. Wang, “Enhanced spatially constrained remotely sensed imagery classification using a fuzzy local double neighborhood information c -means clustering algorithm,” IEEE Journal of Selected Topics in 13 > Applied Earth Observations and...

    2018

  15. [23]

    A new sparse subspace clustering algorithm for hyperspectral remote sensing imagery,

    H. Zhai, H. Zhang, L. Zhang, P. Li, and A. Plaza, “A new sparse subspace clustering algorithm for hyperspectral remote sensing imagery,” IEEE Geoscience and Remote Sensing Letters, vol. 14, no. 1, pp. 43-47, 2016

    2016

  16. [24]

    Scalable graph -based clustering with nonnegative relaxation for large hyperspectral image,

    R. Wang, F. Nie, Z. Wang, F. He, and X. Li, “Scalable graph -based clustering with nonnegative relaxation for large hyperspectral image,” IEEE Transactions on Geoscience Remote Sensing, vol. 57, no. 10, pp. 7352-7364, 2019

    2019

  17. [25]

    Multi -feature fusion: Graph neural network and CNN combining for hyperspectral image classification,

    Y. Ding, Z. Zhang, X. Zhao, D. Hong, W. Cai, C. Yu, N. Yang, and W. Cai, “Multi -feature fusion: Graph neural network and CNN combining for hyperspectral image classification,” Neurocomputing, vol. 501, pp. 246-257, 2022

    2022

  18. [26]

    Spectral–spatial feature extraction with dual graph autoencoder for hyperspectral image clustering,

    Y. Zhang, Y. Wang, X. Chen, X. Jiang, and Y. Zhou, “Spectral–spatial feature extraction with dual graph autoencoder for hyperspectral image clustering,” IEEE Transactions on Circuits and Systems for Video Technology, vol. 32, no. 12, pp. 8500-8511, 2022

    2022

  19. [27]

    Self-supervised locality preserving low- pass graph convolutional embedding for large -scale hyperspectral image clustering,

    Y. Ding, Z. Zhang, X. Zhao, Y. Cai, S. Li, B. Deng, and W. Cai, “Self-supervised locality preserving low- pass graph convolutional embedding for large -scale hyperspectral image clustering,” IEEE Transactions on Geoscience and Remote Sensing, vol. 60, pp. 1-16, 2022

    2022

  20. [28]

    SDST: Self -Supervised Double -Structure Transformer for Hyperspectral Images Clustering,

    F. Luo, Y. Liu, Y. Duan, T. Guo, L. Zhang, and B. Du, “SDST: Self -Supervised Double -Structure Transformer for Hyperspectral Images Clustering,” IEEE Transactions on Geoscience and Remote Sensing, 2024

    2024

  21. [29]

    Attributed graph clustering: A deep attentional embedding approach,

    C. Wang, S. Pan, R. Hu, G. Long, J. Jiang, and C. Zhang, “Attributed graph clustering: A deep attentional embedding approach,” arXiv preprint arXiv:1906.06532, 2019

    1906 arXiv

  22. [30]

    Multiview clustering via hypergraph induced semi - supervised symmetric nonnegative matrix factorization,

    S. Peng, J. Yin, Z. Yang, B. Chen, and Z. Lin, “Multiview clustering via hypergraph induced semi - supervised symmetric nonnegative matrix factorization,” IEEE Transactions on Circuits and Systems for Video Technology, vol. 33, no. 10, pp. 5510-5524, 2023

    2023

  23. [31]

    Local aggregation and global attention network for hyperspectral image classification with spectral-induced aligned superpixel segmentation,

    Z. Chen, G. Wu, H. Gao, Y. Ding, D. Hong, and B. Zhang, “Local aggregation and global attention network for hyperspectral image classification with spectral-induced aligned superpixel segmentation,” Expert systems with applications, vol. 232, pp. 120828, 2023

    2023

  24. [32]

    Homophily -enhanced structure learning for graph clustering

    M. Gu, G. Yang, S. Zhou, N. Ma, J. Chen, Q. Tan, M. Liu, and J. Bu, "Homophily -enhanced structure learning for graph clustering." pp. 577-586

  25. [33]

    Homophily -Related: Adaptive Hybrid Graph Filter for Multi -View Graph Clustering

    Z. Wen, Y. Ling, Y. Ren, T. Wu, J. Chen, X. Pu, Z. Hao, and L. He, "Homophily -Related: Adaptive Hybrid Graph Filter for Multi -View Graph Clustering." pp. 15841-15849

  26. [34]

    Contrastive learning under heterophily,

    W. Yang, and B. Mirzasoleiman, “Contrastive learning under heterophily,” arXiv preprint arXiv:2303.06344, 2023

    2023 arXiv

  27. [35]

    Beyond homophily: Reconstructing structure for graph -agnostic clustering

    E. Pan, and Z. Kang, "Beyond homophily: Reconstructing structure for graph -agnostic clustering." pp. 26868-26877

  28. [36]

    Simple contrastive graph clustering,

    Y. Liu, X. Yang, S. Zhou, X. Liu, S. Wang, K. Liang, W. Tu, and L. Li, “Simple contrastive graph clustering,” IEEE Transactions on Neural Networks and Learning Systems, 2023

    2023

  29. [37]

    SLIC superpixels compared to state-of-the-art superpixel methods,

    R. Achanta, A. Shaji, K. Smith, A. Lucchi, P. Fua, and S. Süsstrunk, “SLIC superpixels compared to state-of-the-art superpixel methods,” IEEE Transactions on Pattern Analysis Machine Intelligence, vol. 34, no. 11, pp. 2274-2282, 2012

    2012

  30. [38]

    Deep fusion clustering network

    W. Tu, S. Zhou, X. Liu, X. Guo, Z. Cai, E. Zhu, and J. Cheng, "Deep fusion clustering network." pp. 9978 - 9987

  31. [39]

    Structural deep clustering network

    D. Bo, X. Wang, C. Shi, M. Zhu, E. Lu, and P. Cui, "Structural deep clustering network." pp. 1400-1410

  32. [40]

    Embedding graph auto -encoder for graph clustering,

    H. Zhang, P. Li, R. Zhang, and X. Li, “Embedding graph auto -encoder for graph clustering,” IEEE Transactions on Neural Networks and Learning Systems, vol. 34, no. 11, pp. 9352-9362, 2022

    2022

  33. [41]

    Adaptive graph auto-encoder for general data clustering,

    X. Li, H. Zhang, and R. Zhang, “Adaptive graph auto-encoder for general data clustering,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 44, no. 12, pp. 9725-9732, 2021

    2021

  34. [42]

    Visualizing data using t-SNE,

    L. Van der Maaten, and G. Hinton, “Visualizing data using t-SNE,” Journal of machine learning research, vol. 9, no. 11, 2008. Yao Ding received the M.S. and Ph.D. degree from the Key Laboratory of Optical Engineering, Xi’an Research Institute of High Technology, Xi’an 710025, ...

    2008

  35. [2022]

    He has published several papers in IEEE Trans

    His research interests include neural network, computer vision, image processing, and hyperspectral image clustering. He has published several papers in IEEE Trans. on Geoscience and Remote Sensing (TGRS), Information Sciences (INS), Expert Systems with Applications (ESWA), De...

  36. [2023]

    At present, he has Eleven highly cited papers of ESI

    He also has received HIGHLY CITED AWARDS from Defence Technology (DT) journal. At present, he has Eleven highly cited papers of ESI. In addition, he is also the reviewer of TGRS, TNNLS, PR, JAG, KBS, etc. He has also served as a Youth editorial board member of Journal of Infor...

    2017

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.