Pith. sign in

REVIEW 5 major objections 3 minor 1 cited by

Hyperspectral Image Classification via Transformer-based Spectral-Spatial Attention Decoupling and Adaptive Gating

T0 review · 5 major / 3 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read STNet explicitly decouples spatial and spectral attention to achieve leading accuracy on three hyperspectral benchmarks.

desk verdict A coherent architecture buried under an internally inconsistent experimental section; the reported numbers cannot be trusted as presented. read the letter →

arxiv 2506.08324 v2 pith:MNM7A6QF submitted 2025-06-10 cs.CV

Guandong Li , Mengxia Ye This is my paper
classification cs.CV
keywords hyperspectralimageclassificationspectral-spatialattentiondecouplingadaptivegatingtransformer3D-DenseNetremotesensing
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

Hyperspectral images pack hundreds of spectral bands per pixel, and classifiers must exploit both spatial context and spectral correlations without drowning in redundancy. STNet, the architecture proposed here, embeds a Transformer module in a 3D-DenseNet that splits attention into a spatial branch and a spectral branch, then adaptively gates how the two are fused and how the feed-forward network transforms the result. The paper claims this explicit decoupling plus dual gating captures discriminative joint spatial-spectral features better than standard 3D-CNNs and transformer baselines, raising overall accuracy to 99.77% on Indian Pines, 100% on Pavia University, and 99.95% on KSC in their experiments. The design is offered as a way to reduce overfitting under small training samples and high noise, which matters because labeled pixels are scarce and expensive in remote sensing. If the claims hold, STNet provides a practical template for handling high-dimensional, spectrally redundant imagery.

What carries the argument

The load-bearing object is the SpatioTemporalTransformer module inserted into a 3D-DenseNet backbone. It carries the argument through four coordinated parts: (1) a spatial attention branch that treats flattened pixel positions as a sequence, modeling intra-band spatial dependencies for all bands simultaneously; (2) a spectral attention branch that reduces spatial dimensions by mean pooling and applies self-attention along the spectral axis, modeling inter-band correlations; (3) an adaptive attention fusion gate that computes a single weight $g$ from the pooled outputs of both branches and fuses them as $g \cdot \mathrm{Attn}_s + (1-g) \cdot \mathrm{Attn}_t$; and (4) a gated feed-forward network, where a parallel linear layer plus sigmoid produces an element-wise gate $\mathrm{g_{ffn}}$ that modulates the main FFN output by Hadamard product. A learnable 3D positional encoding, interpolated trilinearly to any input size, supplies position information. The module's role is to replace the fixed, mixed spatial-spectral processing of 3D convolutions with targeted attention and data-dependent gating, which the paper argues reduces redundant computation and overfitting.

What would settle it

Take the Indian Pines scene, train STNet on a spatially contiguous region, and test on a disjoint region separated by a buffer of unused pixels; if overall accuracy falls well below the reported 99.77%, the claimed generalization is not supported. Alternatively, inspect the released training and test block indices for overlap between the two sets.

Watch

Extended reading notes

Core claim

The central claim is that explicitly separating spatial and spectral attention, rather than letting a single attention mechanism process all tokens, yields better hyperspectral classification. The spatial branch applies multi-head self-attention to the flattened spatial positions of each spectral band, capturing intra-band spatial dependencies; the spectral branch mean-pools spatial dimensions and applies self-attention along the band axis, capturing inter-band spectral correlations. A learned scalar gate $g = \sigma(\mathrm{Linear}_2(\mathrm{ReLU}(\mathrm{Linear}_1([h_s, h_t]))))$ blends the two attention outputs, and a gated feed-forward network multiplies its main output by an element-wise sigmoid gate. On top of this, the paper modifies 3D-DenseNet by exponentially growing the growth rate with block depth and connecting the input to all later layers. The paper reports that these components together outperform SSRN, 3D-CNN, 3D-SE-DenseNet, DGCNet, and Spectralformer on Indian Pines, Pavia University, and KSC, with fast convergence and stable training curves.

Load-bearing premise

The reported accuracies come from randomly splitting pixels of the same scene into train, validation, and test blocks, and because neighboring pixels in a hyperspectral image are highly correlated and blocks may overlap spatially, the test set can leak spatial context from the training set; the paper gives no spatial decontamination procedure.

Editorial extensions

If this is right

  • If the accuracy holds, the decoupled attention design can be transplanted to other 3D data with a redundant axis, such as multi-spectral video or medical volume data.
  • The adaptive fusion gate provides a parameter-light way to balance two attention streams, suggesting that adding gating to existing hybrid CNN-transformer models can improve accuracy without deepening or widening the network.
  • The interpolatable positional encoding means a single trained STNet can be applied to different input block sizes, simplifying deployment across sensors with different spatial resolutions.
  • The claimed reduction in overfitting implies STNet can be trained effectively with fewer labeled pixels per class, which is the typical constraint in real HSI mapping.

Reading between the lines

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

  • The absence of a spatially disjoint evaluation means the headline numbers should be read as within-scene accuracy; a proper test of cross-scene generalization would need block-separated training and test regions, and that test is not reported.
  • The decoupling principle suggests a testable recipe for other modalities: split attention along the axis with the most redundancy, gate the fusion, and measure whether parameter efficiency improves; applying this to video or time-series would show whether the benefit is specific to hyperspectral data.
  • One could extend the single fusion gate to a per-channel or per-head gate, or to more than two attention branches, and compare whether the accuracy gain scales with the number of decoupled streams.
  • The paper's growth-rate and dense-connectivity modifications to DenseNet are presented as secondary, but they may interact with the gating; ablations isolating them would clarify how much of the gain is due to the transformer module versus the backbone changes.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

5 major / 3 minor

Summary. The paper proposes STNet, a hyperspectral image classification network that embeds a SpatioTemporalTransformer module into a 3D-DenseNet backbone. The transformer module explicitly decouples attention into separate spatial and spectral branches, fuses them with a learned adaptive attention fusion gate, and applies a gated feed-forward network for feature selection. The authors also introduce a learnable and interpolatable 3D positional encoding, an exponentially increasing growth rate in dense blocks, and fully dense connectivity. The central claim is that STNet achieves superior classification accuracy on Indian Pines, Pavia University, and KSC compared with mainstream methods such as SSRN, 3D-CNN, 3D-SE-DenseNet, DGCNet, and Spectralformer, while reducing overfitting in small-sample and noisy scenarios. The entire claim rests on the experimental tables in Sections 3.3 and 3.4.

Significance. If the reported results were reproducible and internally consistent, STNet would be a plausible incremental contribution to hyperspectral image classification. The architecture is described in sufficient detail to be reimplemented, and the idea of decoupling spatial and spectral attention with adaptive gating is a reasonable design direction. However, the empirical evidence as presented is not credible: the selected block sizes are inconsistent across sections and tables, Pavia University is reported at exactly 100.00 accuracy on all metrics, hyperparameters are tuned using the test benchmarks, and no code or exact split description is provided. The claimed superiority over mainstream methods is therefore unsupported. The paper contains no machine-checked proofs or reproducibility artifacts, and its only evidence is experimental.

major comments (5)
  1. [Sections 3.2.2, 3.4, Tables 4 and 7] The selected input block size is stated as 15 for Indian Pines and 17 for Pavia/KSC in Section 3.2.2, but Section 3.4 states that all three datasets use 17×17 input patches, and Table 4 leaves block sizes 15 and 17 unmeasured ('-'). In addition, the STNet-base result in Table 7 (OA=99.77, AA=99.66, K=99.74) exactly matches the Table 4 row for block size 13, not the stated selected size. These contradictions make it impossible to determine which configuration produced the headline results.
  2. [Tables 5 and 9] Table 5 reports OA=AA=Kappa=1.000 for block size 17 on Pavia University, and Table 9 reports STNet at exactly 100.00 for OA, AA, and Kappa on the same dataset. With a 4:1:5 train/validation/test split, the test set contains roughly half of all labeled Pavia pixels, and perfect classification of every test pixel across all nine classes is statistically implausible. No standard deviations are reported for STNet in Tables 8–9. These values suggest either test-set contamination through spatially overlapping blocks, a different protocol than described, or misreported numbers.
  3. [Sections 3.2.1 and 3.2.2] The training ratio and block size are selected by comparing OA on the same benchmark test sets in Tables 1–6. This is a form of test-set peeking: the reported accuracies are not independent estimates of generalization. The only validation procedure described is early stopping with the validation portion, which does not justify selecting hyperparameters on the test data. The 'superior performance' claim is therefore partly an artifact of tuning on the test benchmarks.
  4. [Section 3.2.2] The train/validation/test split samples adjacent pixel blocks from the same hyperspectral image without any spatial decontamination, such as disjoint scene regions or buffer zones. Because neighboring HSI pixels are highly correlated and padded blocks can overlap spatially, pixel-level random splits can leak spatial context into the test set and inflate accuracy. The near-perfect class-wise accuracies in Table 8 are consistent with this concern, and the paper provides no analysis showing that test blocks are spatially separated from training blocks.
  5. [Contribution 3 and Section 2] Contribution 3 states that STNet 'improves model representation capability through wavelet convolution with expanded receptive fields,' but the method section never defines or uses a wavelet convolution. The architecture description and Figure 4 contain no wavelet operation. This unfulfilled promise suggests that the text and experiments are not derived from a single coherent implementation, further weakening confidence in the reported results.
minor comments (3)
  1. [Tables 4 and 6] Table 4 and Table 6 leave block sizes 15 and 17 unmeasured on Indian Pines and KSC despite these being the selected sizes according to Section 3.2.2, making the basis for the selection impossible to verify.
  2. [Tables 8 and 9] The comparison method labeled 'Hit' in Tables 8 and 9 is never introduced or described in the text; the reader cannot determine what architecture or configuration it represents.
  3. [Section 2.2.1, Eqs. (1)–(9)] The notation for tensor dimensions is inconsistent, with 'C' used both as the channel dimension and as the model dimension d_model, and the shape of the gate weight g in Eq. (9) is described only vaguely as 'possibly a scalar or a channel-related vector.' Please define all shapes precisely.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity found: STNet's empirical claims are benchmark measurements, not derivations from their own outputs.

full rationale

The paper makes no derivation that reduces to its experimental input or to a self-citation chain. The central claim is an empirical accuracy comparison on Indian Pines, Pavia University, and KSC; the reported OA/AA/Kappa are measured on held-out splits rather than predicted from fitted parameters. No equation in Sections 2.2-2.3 is defined in terms of the experimental results, and no fitted parameter is renamed as a prediction. The many citations to the authors' own prior HSI papers (e.g., DGCNet, LGCNet, WCNet) are contextual related-work references and are not load-bearing for STNet's attention/gating design, which is specified from first principles in the text. The internal inconsistencies noted by the reader (block-size selection vs. Table 7; Contribution 3's 'wavelet convolution' never appearing in methods; near-perfect table entries) are experimental-reporting credibility issues and are not instances of circular reasoning, since none of the reported accuracies is an input to the model definition or to a 'prediction' equation. Per the hard rules, such concerns belong under correctness risk, not circularity; therefore the circularity score is 0.

Assumptions & free parameters 8 free parameters · 5 assumptions · 4 invented entities

The central claim rests on architecture hyperparameters chosen by hand and on an unvalidated evaluation protocol. The paper introduces no physical entities, but it introduces several novel architectural components whose independent benefit is not demonstrated.

free parameters (8)
  • Spatial block size (M,N) = 15 for Indian Pines, 17 for Pavia University and KSC
    Selected by comparing OA in Tables 4 to 6. The selected values are missing from the tables for Indian Pines and KSC.
  • Train/validation/test ratio = 6:1:3 for IN, 4:1:5 for UP, 5:1:4 for KSC
    Chosen from Tables 1 to 3 to balance learning and evaluation reliability. This is not a fixed protocol across datasets.
  • Growth rate k0 = 8, doubling to 16 and 32
    Base of the exponentially increasing growth rate; chosen by design and not justified against alternatives.
  • Gate factor = 0.25
    Reported in the experimental setup without sensitivity analysis.
  • Compression ratio = 16
    Reported in the experimental setup without sensitivity analysis.
  • Number of attention heads = 4
    Reported in the experimental setup without sensitivity analysis.
  • Group convolution groups = 4
    Reported for the 3x3 group convolution without sensitivity analysis.
  • Training epochs = 80
    Used for all datasets. No early-stopping details are given despite mention of validation-based early stopping.
assumptions (5)
  • standard math Multi-head self-attention with learnable projection matrices computes as described in Section 2.2.1.
    Relies on standard Transformer definitions; no proof or implementation is provided.
  • domain assumption Average pooling over spatial dimensions gives representative features for each spectral band (Equation 3).
    This design choice is asserted, not compared with alternatives.
  • domain assumption Trilinear interpolation of a learnable positional encoding preserves position information for arbitrary input sizes (Equation 10).
    Interpolation is a standard function, but its adequacy for HSI blocks is not validated.
  • ad hoc to paper Increasing DenseNet growth rate with depth improves efficiency without adding hyperparameters (Section 2.3.1).
    The paper states this follows from [5], but the formula k=2^(m-1)k0 introduces k0, and no efficiency measurement is provided.
  • ad hoc to paper Random splitting of adjacent pixel blocks into train and test sets is a fair evaluation protocol (Section 3.2.2).
    This is load-bearing and unflagged. Spatial autocorrelation can leak labels between train and test blocks.
invented entities (4)
  • STNet SpatioTemporalTransformer module
    purpose: Replaces mixed spatial-spectral 3D convolutions with decoupled spatial and spectral attention branches, a fusion gate, and a gated feed-forward network.
    The module is only supported by the paper's own tables. No code, weights, or external validation are provided.
  • Adaptive Attention Fusion Gate
    purpose: Learns a scalar or channel weight to balance spatial and spectral attention outputs.
    No standalone evaluation is presented; it is only evaluated as part of STNet.
  • Gated Feed-Forward Network (GFFN)
    purpose: Applies an elementwise gate to the feed-forward output to filter features.
    No standalone evaluation is presented; it is only evaluated as part of STNet.
  • Learnable interpolatable 3D positional encoding
    purpose: Injects position information into the Transformer for 3D HSI cubes.
    Described in Section 2.2.2 with no independent evidence of benefit.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Hyperspectral Image Classification via Transformer-based Spectral-Spatial Attention Decoupling and Adaptive Gating." pith.science (2026). https://pith.science/paper/MNM7A6QF

@misc{pith2026250608324,
  author       = {Pith},
  title        = {Pith review of: Hyperspectral Image Classification via Transformer-based Spectral-Spatial Attention Decoupling and Adaptive Gating},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/MNM7A6QF}},
  note         = {Machine review of arXiv:2506.08324}
}
read the original abstract

Deep neural networks face several challenges in hyperspectral image classification, including high-dimensional data, sparse distribution of ground objects, and spectral redundancy, which often lead to classification overfitting and limited generalization capability. To more effectively extract and fuse spatial context with fine spectral information in hyperspectral image (HSI) classification, this paper proposes a novel network architecture called STNet. The core advantage of STNet stems from the dual innovative design of its Spatial-Spectral Transformer module: first, the fundamental explicit decoupling of spatial and spectral attention ensures targeted capture of key information in HSI; second, two functionally distinct gating mechanisms perform intelligent regulation at both the fusion level of attention flows (adaptive attention fusion gating) and the internal level of feature transformation (GFFN). This characteristic demonstrates superior feature extraction and fusion capabilities compared to traditional convolutional neural networks, while reducing overfitting risks in small-sample and high-noise scenarios. STNet enhances model representation capability without increasing network depth or width. The proposed method demonstrates superior performance on IN, UP, and KSC datasets, outperforming mainstream hyperspectral image classification approaches.

Figures

Figures reproduced from arXiv: 2506.08324 by the authors.

Figure 1
Figure 1. Left: Standard self-attention structure; Right: Our SpatioTemporalTrans [PITH_FULL_IMAGE:figures/full_fig_p006_1.png] view at source ↗
Figure 2
Figure 2. Design of SpatioTemporalTransformer in 3D-DenseNet’s dense block [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗
Figure 3
Figure 3. Proposed DenseNet variant with two key differences from original DenseNet: [PITH_FULL_IMAGE:figures/full_fig_p011_3.png] view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: : Overall architecture of our STNet, incorporating the 3D-DenseNet basic [PITH_FULL_IMAGE:figures/full_fig_p012_4.png]
Figure 5
Figure 5. Figure 5: : False color composite and ground truth labels of Indian Pines dataset [PITH_FULL_IMAGE:figures/full_fig_p013_5.png]
Figure 6
Figure 6. Figure 6: : False color composite and ground truth labels of Pavia University dataset [PITH_FULL_IMAGE:figures/full_fig_p013_6.png]
Figure 7
Figure 7. Figure 7: : Classification results comparison for Indian Pines dataset: (a) False color [PITH_FULL_IMAGE:figures/full_fig_p017_7.png]
Figure 8
Figure 8. Figure 8: : Training and validation curves of STNet showing loss and accuracy evolution [PITH_FULL_IMAGE:figures/full_fig_p018_8.png]

Discussion (0). Sign in to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. MVNet: Hyperspectral Remote Sensing Image Classification Based on Hybrid Mamba-Transformer Vision Backbone Architecture

    cs.CV 2025-07 reject novelty 4.0 of 10

    MVNet combines 3D-CNN, Transformer, and Mamba in a dual-branch design and claims 99%+ accuracy on three hyperspectral benchmarks, but the experimental reporting is internally inconsistent and the method is not reproducible.

Reference graph

Works this paper leans on

28 extracted references · 20 canonical work pages · cited by 1 Pith paper

  1. [1]

    Intelligent remote sensing satellite system.Journal of Remote Sensing, 15(3):415–431, 2011

    Zhang Bing. Intelligent remote sensing satellite system.Journal of Remote Sensing, 15(3):415–431, 2011

    work page 2011

  2. [2]

    Springer Science & Business Media, 2003

    Chein-I Chang.Hyperspectral imaging: techniques for spectral detection and classification, volume 1. Springer Science & Business Media, 2003

    work page 2003

  3. [3]

    Yushi Chen, Hanlu Jiang, Chunyang Li, Xiuping Jia, and Pedram Ghamisi. Deep fea- ture extraction and classification of hyperspectral images based on convolutional neural networks.IEEE transactions on geoscience and remote sensing, 54(10):6232–6251, 2016

    work page 2016

  4. [4]

    Spectralformer: Rethinking hyperspectral image classification with transform- ers.IEEE Transactions on Geoscience and Remote Sensing, 60:1–15, 2021

    Danfeng Hong, Zhu Han, Jing Yao, Lianru Gao, Bing Zhang, Antonio Plaza, and Jocelyn Chanussot. Spectralformer: Rethinking hyperspectral image classification with transform- ers.IEEE Transactions on Geoscience and Remote Sensing, 60:1–15, 2021

    work page 2021

  5. [5]

    Densely connected convolutional networks

    Gao Huang, Zhuang Liu, Laurens Van Der Maaten, and Kilian Q Weinberger. Densely connected convolutional networks. InProceedings of the IEEE conference on computer vision and pattern recognition, pages 4700–4708, 2017

    2017

  6. [6]

    Going deeper with contextual cnn for hyperspectral image classification.IEEE Transactions on Image Processing, 26(10):4843–4855, 2017

    Hyungtae Lee and Heesung Kwon. Going deeper with contextual cnn for hyperspectral image classification.IEEE Transactions on Image Processing, 26(10):4843–4855, 2017

    work page 2017

  7. [7]

    DGCNet: An Efficient 3D-Densenet based on Dynamic Group Convolution for Hyperspectral Remote Sensing Image Classification

    Guandong Li. Dgcnet: An efficient 3d-densenet based on dynamic group convolution for hyperspectral remote sensing image classification.arXiv preprint arXiv:2307.06667, 2023

    work page Pith review arXiv 2023

  8. [8]

    3d wavelet convolutions with extended receptive fields for hyperspectral image classification.arXiv preprint arXiv:2504.10795, 2025

    Guandong Li and Mengxia Ye. 3d wavelet convolutions with extended receptive fields for hyperspectral image classification.arXiv preprint arXiv:2504.10795, 2025

    arXiv 2025

Show all 28 references
  1. [9]

    Efficient dynamic attention 3d convolution for hyperspec- tral image classification.arXiv preprint arXiv:2503.23472, 2025

    Guandong Li and Mengxia Ye. Efficient dynamic attention 3d convolution for hyperspec- tral image classification.arXiv preprint arXiv:2503.23472, 2025

    2025 arXiv

  2. [10]

    Expert kernel generation network driven by contextual mapping for hyperspectral image classification.arXiv preprint arXiv:2504.13045, 2025

    Guandong Li and Mengxia Ye. Expert kernel generation network driven by contextual mapping for hyperspectral image classification.arXiv preprint arXiv:2504.13045, 2025

    2025 arXiv

  3. [11]

    Spatial-geometry enhanced 3d dynamic snake con- volutional neural network for hyperspectral image classification.arXiv preprint arXiv:2504.04463, 2025

    Guandong Li and Mengxia Ye. Spatial-geometry enhanced 3d dynamic snake con- volutional neural network for hyperspectral image classification.arXiv preprint arXiv:2504.04463, 2025

    2025 arXiv

  4. [12]

    Spatial-spectral hyperspectral classification based on learnable 3d group convolution.Spectroscopy Letters, pages 1–13, 2025

    Guandong Li and Mengxia Ye. Spatial-spectral hyperspectral classification based on learnable 3d group convolution.Spectroscopy Letters, pages 1–13, 2025

    2025

  5. [13]

    Faster hyperspectral image classification based on selective kernel mechanism using deep convolutional networks.arXiv preprint arXiv:2202.06458, 2022

    Guandong Li and Chunju Zhang. Faster hyperspectral image classification based on selective kernel mechanism using deep convolutional networks.arXiv preprint arXiv:2202.06458, 2022

    2022 arXiv

  6. [14]

    Doubleconvpool-structured 3d-cnn for hyper- spectral remote sensing image classification.Journal of Image and Graphics, 24(4):639– 654, 2019

    GD Li, CJ Zhang, F Gao, and XY Zhang. Doubleconvpool-structured 3d-cnn for hyper- spectral remote sensing image classification.Journal of Image and Graphics, 24(4):639– 654, 2019

    2019

  7. [15]

    Hyperspectral remote sensing image classification using three-dimensional-squeeze-and- excitation-densenet (3d-se-densenet).Remote Sensing Letters, 11(2):195–203, 2020

    Guandong Li, Chunju Zhang, Runmin Lei, Xueying Zhang, Zhourun Ye, and Xiaoli Li. Hyperspectral remote sensing image classification using three-dimensional-squeeze-and- excitation-densenet (3d-se-densenet).Remote Sensing Letters, 11(2):195–203, 2020

    2020

  8. [16]

    Scene classi- fication of high-resolution remote sensing image using transfer learning with multi-model feature extraction framework

    Guandong Li, Chunju Zhang, Mingkai Wang, Fei Gao, and Xueying Zhang. Scene classi- fication of high-resolution remote sensing image using transfer learning with multi-model feature extraction framework. InImage and Graphics Technologies and Applications: 13th Conference on Ima...

    2018

  9. [17]

    Dual classification head self-training network for cross-scene hyperspectral image classification.arXiv preprint arXiv:2502.17879, 2025

    Rong Liu, Junye Liang, Jiaqi Yang, Jiang He, and Peng Zhu. Dual classification head self-training network for cross-scene hyperspectral image classification.arXiv preprint arXiv:2502.17879, 2025

    2025 arXiv

  10. [18]

    Learning efficient convolutional networks through network slimming

    Zhuang Liu, Jianguo Li, Zhiqiang Shen, Gao Huang, Shoumeng Yan, and Changshui Zhang. Learning efficient convolutional networks through network slimming. InProceed- ings of the IEEE international conference on computer vision, pages 2736–2744, 2017

    2017

  11. [19]

    Swin transformer: Hierarchical vision transformer using shifted windows

    Ze Liu, Yutong Lin, Yue Cao, Han Hu, Yixuan Wei, Zheng Zhang, Stephen Lin, and Baining Guo. Swin transformer: Hierarchical vision transformer using shifted windows. In Proceedings of the IEEE/CVF international conference on computer vision, pages 10012– 10022, 2021

    2021

  12. [20]

    Kan: Kolmogorov-arnold networks.arXiv preprint arXiv:2404.19756, 2024

    Ziming Liu, Yixuan Wang, Sachin Vaidya, Fabian Ruehle, James Halverson, Marin Soljaˇ ci´ c, Thomas Y Hou, and Max Tegmark. Kan: Kolmogorov-arnold networks.arXiv preprint arXiv:2404.19756, 2024

    2024 arXiv

  13. [21]

    Deep supervised learning for hyperspectral data classification through con- volutional neural networks

    Konstantinos Makantasis, Konstantinos Karantzalos, Anastasios Doulamis, and Nikolaos Doulamis. Deep supervised learning for hyperspectral data classification through con- volutional neural networks. In2015 IEEE international geoscience and remote sensing 19 symposium (IGARSS),...

    2015

  14. [22]

    Spectral–spatial feature tokeniza- tion transformer for hyperspectral image classification.IEEE Transactions on Geoscience and Remote Sensing, 60:1–14, 2022

    Le Sun, Guangrui Zhao, Yuhui Zheng, and Zebin Wu. Spectral–spatial feature tokeniza- tion transformer for hyperspectral image classification.IEEE Transactions on Geoscience and Remote Sensing, 60:1–14, 2022

    2022

  15. [23]

    Kolmogorov- arnold networks (kans) for time series analysis.arXiv preprint arXiv:2405.08790, 2024

    Cristian J Vaca-Rubio, Luis Blanco, Roberto Pereira, and M` arius Caus. Kolmogorov- arnold networks (kans) for time series analysis.arXiv preprint arXiv:2405.08790, 2024

    2024

  16. [24]

    A fast dense spectral– spatial convolution network framework for hyperspectral images classification.Remote sensing, 10(7):1068, 2018

    Wenju Wang, Shuguang Dou, Zhongmin Jiang, and Liujie Sun. A fast dense spectral– spatial convolution network framework for hyperspectral images classification.Remote sensing, 10(7):1068, 2018

    2018

  17. [25]

    Multi-scale dense networks for hyperspec- tral remote sensing image classification.IEEE Transactions on Geoscience and Remote Sensing, 57(11):9201–9222, 2019

    Chunju Zhang, Guandong Li, and Shihong Du. Multi-scale dense networks for hyperspec- tral remote sensing image classification.IEEE Transactions on Geoscience and Remote Sensing, 57(11):9201–9222, 2019

    2019

  18. [26]

    Three-dimensional densely connected convolutional network for hyperspectral remote sensing image classifi- cation.Journal of Applied Remote Sensing, 13(1):016519–016519, 2019

    Chunju Zhang, Guandong Li, Shihong Du, Wuzhou Tan, and Fei Gao. Three-dimensional densely connected convolutional network for hyperspectral remote sensing image classifi- cation.Journal of Applied Remote Sensing, 13(1):016519–016519, 2019

    2019

  19. [27]

    Chunju Zhang, Guandong Li, Runmin Lei, Shihong Du, Xueying Zhang, Hui Zheng, and Zhaofu Wu. Deep feature aggregation network for hyperspectral remote sensing image classification.IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 13:5314–5325, 2020

    2020

  20. [28]

    Wenzhi Zhao and Shihong Du. Spectral–spatial feature extraction for hyperspectral image classification: A dimension reduction and deep learning approach.IEEE Transactions on Geoscience and Remote Sensing, 54(8):4544–4554, 2016. 20

    2016

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.