Pith. sign in

REVIEW 4 major objections 4 minor 48 references

GeloVec: Higher Dimensional Geometric Smoothing for Coherent Visual Feature Extraction in Image Segmentation

T0 review · 4 major / 4 minor · reviewed 2026-08-16 · deepseek-v4-flash

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

Pith's one-line read GeloVec claims geometric smoothing via Chebyshev distances and an orthogonal basis transform stabilizes attention and lifts segmentation accuracy on three benchmarks.

desk verdict GeloVec's headline gains don't survive contact with its own equations: the geometric attention modulation is a softmax no-op, and the tables contradict the abstract. read the letter →

arxiv 2505.01057 v1 pith:NX3XPZ6H submitted 2025-05-02 cs.CV

Boris Kriuk , Matey Yordanov This is my paper · ORCID
classification cs.CV MSC 68U1068T4515A6653A0514R0568T07
keywords semanticsegmentationgeometricsmoothingChebyshevdistanceattentionmechanismorthogonalbasistransformfeatureextractionUNetResNet-34
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

GeloVec is a proposal for treating attention in semantic segmentation as a geometric operation: instead of letting attention scores drift near object boundaries, it smooths features by computing Chebyshev distances in a higher-dimensional orthogonal feature space and uses those distances to modulate attention and preserve edges. The paper claims this geometric smoothing stabilizes attention, improves boundary precision, and keeps intra-class regions homogeneous, and reports that a UNet with a ResNet-34 backbone equipped with GeloVec modules beats U-Net, DeepLabV3+, HRNet, and SegFormer on Caltech Birds-200, LSDSC, and FSSD, with mean IoU gains of 2.1, 2.7, and 2.4 percentage points respectively. If the reported numbers hold, the practical significance is that a lightweight geometric block added between encoder stages can improve segmentation without switching to a heavier backbone or a transformer.

What carries the argument

The load-bearing object is the GeloVec module, whose steps are given in equations (3.1)-(3.6): an Orthogonal Basis Transform (OBT) projects features into a normalized higher-dimensional orthogonal basis; Geometric Adaptive Sampling (GAS) computes the Chebyshev distance $D_{\mathrm{chebyshev}}(p_c,p_i)=\max_{d}|W_i(F_{p_c,d}-F_{p_i,d})|$ between a center pixel and weighted dilated neighbors; a $1\times1$ convolution plus sigmoid converts the max over neighbors into a normalized distance $D_{\mathrm{norm}}$; and an edge gate $G_{\mathrm{edge}}=\sigma(\mathrm{conv}(D_{\mathrm{norm}}))$ blends original and edge features. The attention logits subtract $\lambda D_{\mathrm{norm}}$ before softmax. The Chebyshev max over channels is what gives the module its boundary sensitivity: a single channel jumping at an edge raises the distance and suppresses attention there, while flat regions keep small distances and stay homogeneous.

What would settle it

Run GeloVec and the four baselines on the same three datasets with shared training settings and multiple seeds, then compare mean IoU; if GeloVec does not beat SegFormer by the reported margins, or if the numbers require unpublished configuration choices to line up, the central claim fails.

Watch

Extended reading notes

Core claim

The central claim is that replacing plain attention with a geometry-aware smoothing module yields more coherent segmentations. GeloVec inserts a module after each encoder stage of a UNet; the Orthogonal Basis Transform projects features into a normalized higher-dimensional orthogonal basis, and Geometric Adaptive Sampling compares each pixel with dilated neighbors using a modified Chebyshev distance, taking the maximum absolute per-channel difference. The normalized distance gates an edge-preserving blend and is subtracted from the scaled dot-product attention logits, so locations with large geometric change are treated as boundaries rather than being smoothed over. The paper reports mean IoU gains of 2.1, 2.7, and 2.4 percentage points over state-of-the-art methods on CUB-200-2011, LSDSC, and FSSD, with table IoU values of 83.6, 85.4, and 82.9 and the largest advantages in precision.

Load-bearing premise

The central comparison assumes the reported IoU numbers were produced by the described architecture under a consistent, reproducible protocol, yet no optimizer, epochs, splits, or run count are given, and the abstract, text, and tables give conflicting values such as 84.3 versus 85.4 for LSDSC.

Editorial extensions

If this is right

  • On the three datasets studied, GeloVec reports higher IoU, F1, precision, and recall than every baseline listed, so the geometric module is claimed to improve all four metrics at once.
  • The precision values (92.1, 90.7, and 89.6) are the largest gaps over the baselines, which the paper attributes to the geometric distance reducing false positives at boundaries.
  • GeloVec achieves these numbers with a ResNet-34 backbone, lighter than DeepLabV3+'s ResNet-50 and HRNet's HRNetV2-W18, implying the gain comes from the module rather than a larger encoder.
  • Because the module sits between encoder stages and its refined features travel through skip connections, the geometric smoothing affects both the semantic and the detail path of the UNet.
  • The abstract claims the transformations are parallelizable and lossless, so the approach is presented as adding accuracy without a meaningful compute penalty.

Reading between the lines

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

  • If the reported gains reproduce, the same geometric smoothing block could be inserted into transformer-based segmentation heads, where boundary instability is also reported; the paper does not test that configuration.
  • A component ablation, removing the edge gate, the Chebyshev modulation, or the orthogonal transform one at a time, would isolate how much of the improvement is boundary preservation versus intra-class smoothing.
  • The normalized distance map $D_{\mathrm{norm}}$ could be exported as an interpretable edge prior for other dense-prediction tasks such as depth estimation or contour detection, an application the paper does not explore.
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

4 major / 4 minor

Summary. The paper proposes GeloVec, a module inserted at multiple encoder stages of a U-Net with ResNet-34 backbone for semantic segmentation. The module combines an orthogonal basis transform, a Chebyshev-distance geometric adaptive sampling term, an edge gate, and a softmax attention mechanism modulated by a geometric distance. The authors report mIoU gains over SegFormer on three datasets (CUB-200-2011, LSDSC, FSSD) and claim that Riemannian geometry provides theoretical guarantees on segmentation stability. The central claim is that geometric smoothing of attention improves boundary preservation and intra-class coherence.

Significance. If the proposed mechanism worked as described, a geometry-based attention smoothing layer that is computationally cheap and pluggable into CNN segmentation backbones would be a useful contribution. The paper is also transparent in giving explicit equations for the proposed modules and in reporting numbers for four baseline architectures, which is commendable. However, the central mechanism is mathematically inert as written: the Chebyshev term added inside the softmax is cancelled by softmax shift invariance. In addition, the experimental numbers contradict each other between the abstract, the text, and the tables, and no training protocol or statistical evaluation is provided. These issues affect the central claims directly, not just the presentation.

major comments (4)
  1. [Sec. 3.2, Eq. (3.6)] The geometric modulation of the attention weights is a no-op. D_norm in Eq. (3.4) is, for each spatial query position, a single scalar: it is the maximum of the local Chebyshev distances over neighbors, passed through a 1x1 convolution and a sigmoid at that spatial location. In Eq. (3.6), this scalar is subtracted from every key logit in the row of the attention matrix for that query. Since softmax is invariant to adding a constant to all logits in a row, subtracting lambda times D_norm changes no attention probability. Therefore, the claimed Chebyshev-distance geometric smoothing of the attention pathway cannot work as described; any observed benefit would have to come from the separate edge gate in Eq. (3.5), which is distinct from the paper's stated core contribution. This is a load-bearing internal inconsistency.
  2. [Abstract vs. Sec. 4, Tables 1-3] The reported numbers are internally contradictory. The abstract and Section 1 claim mIoU gains of 2.1%, 2.7%, and 2.4% over state-of-the-art methods on CUB-200-2011, LSDSC, and FSSD, respectively. Table 1 shows GeloVec at 83.6 vs. SegFormer at 83.5, a gain of only 0.1 points on CUB-200-2011. Table 3 shows 82.9 vs. 81.2, a gain of 1.7 points on FSSD. Additionally, Section 4 states an LSDSC IoU of 84.3% for GeloVec, while Table 2 lists 85.4%. These discrepancies mean the abstract's headline figures are not reproducible from the tables in the same manuscript, and the high-level claim of consistent 2+ point gains is unsupported.
  3. [Sec. 4, Experimental protocol] The evaluation section provides no training hyperparameters, optimizer, learning rate schedule, batch size, number of epochs, loss function, data split details, or number of runs for any method. No error bars or statistical significance tests are reported. This is insufficient for a claim of consistent superiority over state-of-the-art baselines, especially when the reported differences are as small as 0.1 mIoU. The absence of ablations is also critical: the paper never isolates the contributions of the orthogonal basis transform, the geometric adaptive sampling, the edge gate, and the attention modulation, so the attribution of any observed gains to the proposed 'geometric smoothing' is not established even if the numbers were reliable.
  4. [Secs. 3 and 5, Theoretical claims] The paper repeatedly asserts that GeloVec has a 'mathematical foundation in Riemannian geometry' and provides 'theoretical guarantees on segmentation stability,' but no Riemannian metric, manifold structure, or stability theorem is defined or proved anywhere. The only geometric elements are the Chebyshev distance in Eq. (3.3) and the L2 normalization in Eq. (3.2). Moreover, Eq. (3.2) calls the reshaped normalized vectors an 'orthogonal basis,' but L2 normalization does not by itself make a set of vectors mutually orthogonal. These statements are thus assertion rather than derivation, and the claimed guarantees are not supported by the presented mathematics.
minor comments (4)
  1. [Throughout] There are recurring typos such as 'spacial' for 'spatial' (e.g., Section 3.1 and Section 5), and the phrase 'L-norm' in Section 3.2 should be 'L-infinity norm' or 'Chebyshev norm' for clarity.
  2. [Eq. (3.3)] The notation W_i for the learnable sampling weight is ambiguous: it is unclear whether the same weight vector is used across all feature channels d, and how the product W_i*(F_pc,d - F_pi,d) is computed when W_i is a scalar, vector, or tensor. The equation and surrounding text should specify the exact dimensions.
  3. [Fig. 1] The architecture overview figure is referenced but not discussed in enough detail in the text; in particular, the placement of the four GeloVec variants (GeloVecLow, Mid, High, VeryHigh) is described verbally but not marked in the figure, which makes the multi-scale integration harder to follow.
  4. [Sec. 4] The visual results in Figures 2 and 3 are presented as qualitative evidence, but no explanation is given for what is being displayed in Figure 2 beyond 'attention values distribution,' and the caption of Figure 3 does not specify which dataset or which images are shown. This limits the usefulness of the qualitative comparison.

Circularity Check

1 steps flagged · score 6.0 of 10

Equation (3.6)'s geometric modulation cancels by softmax shift invariance; the central attention-smoothing mechanism is a no-op by construction.

  1. other [Section 3.2, Equations (3.3)-(3.6)]
    "Dnorm =σ (Conv1× 1 (maxi∈N (pc)Dchebyshev(pc,p i))) (3.4) ... Araw = Softmax(QK T /√dk −λ·Dnorm) (3.6) ... attention scores are computed and modulated by the geometric distance metric."

    By equation (3.4), Dnorm is a single per-pixel scalar at query position pc. In equation (3.6), for a fixed pc, the row of QK^T contains one logit per key, and subtracting λ·Dnorm(pc) from every entry in that row is a row-constant shift. Softmax is shift-invariant: Softmax(x+c)=Softmax(x). Therefore A_raw = Softmax(QK^T/√dk) identically; the Chebyshev distance term cancels by construction and does not modulate attention at all. The paper's central claim that the geometric distance 'modulates' attention and stabilizes it reduces, through its own equations, to standard unmodulated self-attention. Any observed benefit would have to come from the separate edge gate in equation (3.5), not from the named contribution.

full rationale

The paper's empirical comparisons against external baselines (Tables 1-3) are independent and not circular, assuming the numbers are reproducible. However, the central theoretical mechanism is circular in a mathematical sense: equation (3.6) subtracts a per-pixel scalar Dnorm inside the softmax, and softmax shift invariance makes that term a no-op. The claimed geometric attention smoothing is therefore equivalent, by the paper's own definitions, to unmodulated self-attention. Separately, the abstract's assertion that 'Riemannian geometry provides theoretical guarantees on segmentation stability' is an omitted proof, not a derived result; no metric or stability theorem is stated anywhere. The text also contains contradictory numbers (e.g., abstract mIoU gains of 2.1/2.4 versus Table 1's 83.6 vs 83.5 and Table 3's 82.9 vs 81.2, and LSDSC text 84.3 versus Table 2's 85.4), but those are reproducibility/correctness issues rather than circularity. No self-citations are load-bearing. Because the paper's flagship contribution—geometric modulation of attention—reduces to identity by construction, the score is 6.

Assumptions & free parameters 4 free parameters · 4 assumptions · 0 invented entities

The central contribution is an architecture whose main ingredients are learned transformations; the paper adds no new physical or mathematical entity. However, it relies on several unsupported modeling assumptions and omits the values of its key hyperparameters.

free parameters (4)
  • Lambda (lambda in Eq. 3.6) = Not specified
    Scales the geometric distance penalty in the attention modulation; chosen by hand, no value reported.
  • GAS neighborhood size and dilation rate = Not specified
    Defines the local window for Chebyshev distance sampling; no value given in Section 3.2.
  • Number of basis vectors n in OBT = Not specified
    Controls the expanded feature dimension in Eq. 3.1-3.2; chosen without stated justification.
  • Learnable sampling weights W_i (Eq. 3.3) = Learned during training
    The 'adaptive sampling weights system' is trained end-to-end on the segmentation objective, so the distance metric is fitted to the data.
assumptions (4)
  • ad hoc to paper Feature maps can be represented as points on a Riemannian manifold where Chebyshev distance approximates perceptual boundary strength
    Invoked in Section 3.2 GAS; no manifold structure, metric tensor, or geodesic is defined, so this is an unsupported modeling assumption.
  • ad hoc to paper L2 normalization of reshaped feature vectors produces an orthogonal basis
    Eq. 3.2 normalizes each vector but does not enforce orthogonality between components; calling the result 'orthogonal' is mathematically inaccurate.
  • domain assumption Geometric smoothing of attention features improves edge preservation and intra-class homogeneity without side effects
    Central design assumption behind the module; no ablation or analysis isolates this effect.
  • domain assumption The three selected datasets are sufficient to establish state-of-the-art performance
    Evaluation uses CUB-200-2011, LSDSC, and FSSD only, with no standard benchmarks (Pascal VOC, Cityscapes, ADE20K) and no comparison to current SOTA models.

how reviews work

0 comments
Cite this review

Pith. "Pith review of GeloVec: Higher Dimensional Geometric Smoothing for Coherent Visual Feature Extraction in Image Segmentation." pith.science (2026). https://pith.science/paper/NX3XPZ6H

@misc{pith2026250501057,
  author       = {Pith},
  title        = {Pith review of: GeloVec: Higher Dimensional Geometric Smoothing for Coherent Visual Feature Extraction in Image Segmentation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/NX3XPZ6H}},
  note         = {Machine review of arXiv:2505.01057}
}
read the original abstract

This paper introduces GeloVec, a new CNN-based attention smoothing framework for semantic segmentation that addresses critical limitations in conventional approaches. While existing attention-backed segmentation methods suffer from boundary instability and contextual discontinuities during feature mapping, our framework implements a higher-dimensional geometric smoothing method to establish a robust manifold relationships between visually coherent regions. GeloVec combines modified Chebyshev distance metrics with multispatial transformations to enhance segmentation accuracy through stabilized feature extraction. The core innovation lies in the adaptive sampling weights system that calculates geometric distances in n-dimensional feature space, achieving superior edge preservation while maintaining intra-class homogeneity. The multispatial transformation matrix incorporates tensorial projections with orthogonal basis vectors, creating more discriminative feature representations without sacrificing computational efficiency. Experimental validation across multiple benchmark datasets demonstrates significant improvements in segmentation performance, with mean Intersection over Union (mIoU) gains of 2.1%, 2.7%, and 2.4% on Caltech Birds-200, LSDSC, and FSSD datasets respectively compared to state-of-the-art methods. GeloVec's mathematical foundation in Riemannian geometry provides theoretical guarantees on segmentation stability. Importantly, our framework maintains computational efficiency through parallelized implementation of geodesic transformations and exhibits strong generalization capabilities across disciplines due to the absence of information loss during transformations.

Figures

Figures reproduced from arXiv: 2505.01057 by the authors.

Figure 1
Figure 1. Architecture Overview. 2.5. Multi-Scale and Multi-Spatial Feature Representation Handling variations in object scale and appearance requires aggregating infor￾mation from different perspectives. Spatial Pyramid Pooling (SPP) [16] and Feature Pyramid Networks (FPN) [22] capture multi-scale context by pooling features at different rates or combining feature maps from different network levels. Multi-view CNNs process d… view at source ↗
Figure 2
Figure 2. Attention Values Distribution Comparison 8 [PITH_FULL_IMAGE:figures/full_fig_p008_2.png] view at source ↗
Figure 3
Figure 3. FSSD Dataset Segmentation Results [PITH_FULL_IMAGE:figures/full_fig_p009_3.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

48 extracted references · 39 canonical work pages

  1. [1]

    Neural ma- chine translation by jointly learning to align and translate

    Dzmitry Bahdanau, Kyunghyun Cho, and Yoshua Bengio. Neural ma- chine translation by jointly learning to align and translate. arXiv preprint arXiv:1409.0473, 2014

    arXiv 2014

  2. [2]

    A Threefold Review on Deep Semantic Segmentation: Efficiency-oriented, Temporal and Depth-aware design

    Felipe M. Barbosa and Fernando S. Os´ orio. A threefold review on deep seman- tic segmentation: Efficiency-oriented, temporal and depth-aware design. arXiv preprint arXiv:2303.04315, 2023

    work page Pith review arXiv 2023

  3. [3]

    Benfenati and A

    A. Benfenati and A. Marta. A singular riemannian geometry approach to deep neural networks i. theoretical foundations.Neural Networks, 158:331–343, 2023

    work page 2023

  4. [4]

    Benfenati and A

    A. Benfenati and A. Marta. A singular riemannian geometry approach to deep neural networks ii. reconstruction of 1-d equivalence classes. Neural networks : the official journal of the International Neural Network Society , 158:344–358, 2023

    work page 2023

  5. [5]

    A. Brandt. Note on the riemannian geometry of image processing. Journal of Mathematical Imaging and Vision , 15(1-2):87–94, 2001

    work page 2001

  6. [6]

    Brauwers and F

    G. Brauwers and F. Frasincar. A general survey on attention mechanisms in deep learning. IEEE Transactions on Knowledge and Data Engineering , 35(4):3279–3298, April 2023

    work page 2023

  7. [7]

    Bronstein, Joan Bruna, Taco Cohen, and Petar Veliˇ ckovi´ c

    Michael M. Bronstein, Joan Bruna, Taco Cohen, and Petar Veliˇ ckovi´ c. Geomet- ric deep learning: Grids, groups, graphs, geodesics, and gauges. arXiv preprint arXiv:2104.13478, 2021

    arXiv 2021

  8. [8]

    Manifold learning of brain mris by deep learning

    Tom Brosch, Roger Tam, and Initiative for the Alzheimers Disease Neuroimag- ing. Manifold learning of brain mris by deep learning. InMedical image comput- ing and computer-assisted intervention : MICCAI ... International Conference on Medical Image Computing and Computer-Assisted Intervention , volume 16, pages 633–640. 2013

    work page 2013

Show all 48 references
  1. [9]

    Cao and Q

    F. Cao and Q. Bao. A survey on image semantic segmentation methods with convolutional neural network. In 2020 International Conference on Communi- cations, Information System and Computer Engineering (CISCE) , pages 458– 462, Kuala Lumpur, Malaysia, 2020

    2020

  2. [10]

    Liang-Chieh Chen, George Papandreou, Iasonas Kokkinos, Kevin Murphy, and Alan L. Yuille. Deeplab: Semantic image segmentation with deep convolutional nets, atrous convolution, and fully connected crfs. IEEE Transactions on Pat- tern Analysis and Machine Intelligence , 40(4):83...

    2018

  3. [11]

    Semantic image seg- mentation: Two decades of research

    Gabriela Csurka, Riccardo Volpi, and Boris Chidlovskii. Semantic image seg- mentation: Two decades of research. Foundations and Trends® in Computer Graphics and Vision , abs/2302.06378(1–2):1–162, 2023

    2023 arXiv

  4. [12]

    Dias and F N S Medeiros

    Felipe M. Dias and F N S Medeiros. Semantic segmentation refinement by monte carlo region growing of high confidence detections. arXiv preprint arXiv:1802.07789, 2018

    2018 arXiv

  5. [13]

    A review on deep learning techniques applied to semantic segmentation

    Alberto Garcia-Garcia, Sergio Orts-Escolano, Sergiu Oprea, Victor Villena- Martinez, and Jose Garcia-Rodriguez. A review on deep learning techniques applied to semantic segmentation. arXiv preprint arXiv:1704.06857 , 2017

    2017 arXiv

  6. [14]

    Martin, Ming-Ming Cheng, and Shi-Min Hu

    Meng-Hao Guo, Tian-Xing Xu, Jiang-Jiang Liu, Zheng-Ning Liu, Peng-Fei Jiang, Tai-Jiang Mu, Shi-Huang Zhang, Ralph R. Martin, Ming-Ming Cheng, and Shi-Min Hu. Attention mechanisms in computer vision: A survey. Compu- tational Visual Media , 8(3):331–368, Sept 2022. 10 Hong Kong...

    2022

  7. [15]

    Principles of riemannian geometry in neural net- works

    Marc Hauser and Asok Ray. Principles of riemannian geometry in neural net- works. In I. Guyon, U. V. Luxburg, S. Bengio, H. Wallach, R. Fergus, S. Vish- wanathan, and R. Garnett, editors, Advances in Neural Information Processing Systems 30 (NIPS 2017) , pages 939–949. Curran ...

    2017

  8. [16]

    Spatial pyramid pooling in deep convolutional networks for visual recognition

    Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Spatial pyramid pooling in deep convolutional networks for visual recognition. In David Fleet, Tomas Pajdla, Bernt Schiele, and Tinne Tuytelaars, editors, Computer Vision – ECCV 2014 , volume 8691 of Lecture Notes in Compu...

    2014

  9. [17]

    Deep manifold learning for dynamic mr imaging

    Ziyu Ke, Wenqi Huang, Jinyuan Cheng, Zheyuan Cui, Sen Jia, Haifeng Wang, Leslie Ying, and Dong Liang. Deep manifold learning for dynamic mr imaging. In Proceedings of the ISMRM & SMRT Annual Meeting & Exhibition , 2021

    2021

  10. [18]

    B. Li, Y. Shi, Z. Qi, and Z. Chen. A survey on semantic segmentation. In 2018 IEEE International Conference on Data Mining Workshops (ICDMW) , pages 1233–1240, Singapore, 2018

    2018

  11. [19]

    Flood semantic segmentation dataset, 2022

    Huayang Li. Flood semantic segmentation dataset, 2022. Accessed: May 1, 2025

    2022

  12. [20]

    Boundary refinement network with semantic embedding connections for uav aerial image semantic segmentation

    Rui Li, Zhong Shi, Fancheng Kong, Xiaobin Zhao, and Tao Luo. Boundary refinement network with semantic embedding connections for uav aerial image semantic segmentation. Journal of Electronic Imaging , 32(06):063003, 2023

    2023

  13. [21]

    G. Lin, A. Milan, C. Shen, and I. Reid. Refinenet: Multi-path refinement net- works for high-resolution semantic segmentation. In 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) , pages 5168–5177, Hon- olulu, HI, USA, 2017

    2017

  14. [22]

    Feature pyramid networks for object detection

    Tsung-Yi Lin, Piotr Doll´ ar, Ross Girshick, Kaiming He, Bharath Hariharan, and Serge Belongie. Feature pyramid networks for object detection. In 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pages 936–944, Honolulu, HI, USA, 2017

    2017

  15. [23]

    Mffnet: A building extraction network for multi-source high-resolution remote sensing data

    Kai Liu, Yuhan Xi, Jiahang Liu, Weichao Zhou, and Yifei Zhang. Mffnet: A building extraction network for multi-source high-resolution remote sensing data. Applied Sciences, 13(24):13067, 2023

    2023

  16. [24]

    Lohit and P

    S. Lohit and P. Turaga. Learning invariant riemannian geometric representa- tions using deep nets. In 2017 IEEE International Conference on Computer Vision Workshops (ICCVW) , pages 1329–1338, Venice, Italy, 2017

    2017

  17. [25]

    J. Long, E. Shelhamer, and T. Darrell. Fully convolutional networks for seman- tic segmentation. In 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pages 3431–3440, Boston, MA, USA, 2015

    2015

  18. [26]

    Boykov, Fatih Porikli, Antonio J

    Shervin Minaee, Yuri Y. Boykov, Fatih Porikli, Antonio J. Plaza, Nasser Ke- htarnavaz, and Demetri Terzopoulos. Image segmentation using deep learning: A survey. IEEE Transactions on Pattern Analysis and Machine Intelligence , 44(7):3523–3542, July 2022

    2022

  19. [27]

    Monti, D

    F. Monti, D. Boscaini, J. Masci, E. Rodol` a, J. Svoboda, and M. M. Bronstein. Geometric deep learning on graphs and manifolds using mixture model cnns. In 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) , pages 5425–5434, Honolulu, HI, USA, 2017

    2017

  20. [28]

    Poincar´ e embeddings for learning hier- archical representations

    Maximillian Nickel and Douwe Kiela. Poincar´ e embeddings for learning hier- archical representations. In I. Guyon, U. V. Luxburg, S. Bengio, H. Wallach, 11 Hong Kong University of Science and Technology R. Fergus, S. Vishwanathan, and R. Garnett, editors, Advances in Neural I...

    2017

  21. [29]

    Contextnet: Exploring context and detail for semantic segmentation in real-time

    R P K Poudel, U D Bonde, S Liwicki, and C Zach. Contextnet: Exploring context and detail for semantic segmentation in real-time. In Richard Bowden, John Collomosse, and Krystian Mikolajczyk, editors, Proceedings of the British Machine Vision Conference (BMVC) . BMVA Press, Sep...

    2018

  22. [30]

    Context-aware semantic segmentation: Enhancing pixel-level understanding with large language models for advanced vision ap- plications

    Borhanuddin Rahman. Context-aware semantic segmentation: Enhancing pixel-level understanding with large language models for advanced vision ap- plications. arXiv preprint arXiv:2503.19276 , 2025

    2025 arXiv

  23. [31]

    U-net: Convolutional networks for biomedical image segmentation, 2015

    Olaf Ronneberger, Philipp Fischer, and Thomas Brox. U-net: Convolutional networks for biomedical image segmentation, 2015

    2015

  24. [32]

    Multi-view classification with convolutional neural networks

    Marco Seeland and Patrick M¨ ader. Multi-view classification with convolutional neural networks. PLoS ONE, 16(1):e0245230, 2021

    2021

  25. [33]

    H. Su, S. Maji, E. Kalogerakis, and E. Learned-Miller. Multi-view convolu- tional neural networks for 3d shape recognition. In 2015 IEEE International Conference on Computer Vision (ICCV), pages 945–953, Santiago, Chile, 2015

    2015

  26. [34]

    C. Tang, H. Chen, X. Li, J. Li, Z. Zhang, and X. Hu. Look closer to seg- ment better: Boundary patch refinement for instance segmentation. In 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 13921–13930, Nashville, TN, USA, 2021

    2021

  27. [35]

    D. V. S. Thisanke, T. G. Chitty, and A. D. A. T. Perera. Semantic segmenta- tion using vision transformers: A survey. Engineering Applications of Artificial Intelligence, 126:106669, 2023

    2023

  28. [36]

    A survey on deep learning-based architec- tures for semantic segmentation on 2d images

    Irem Ulku and Erdem Akag¨ und¨ uz. A survey on deep learning-based architec- tures for semantic segmentation on 2d images. Applied Artificial Intelligence , 36(1), 2022

    2022

  29. [37]

    A large-scale dataset for fish segmentation and classification

    Oguzhan Ulucan, Diclehan Karakaya, and Mehmet Turkan. A large-scale dataset for fish segmentation and classification. pages 1–5, 2020

    2020

  30. [38]

    Gomez, Lukasz Kaiser, and Illia Polosukhin

    Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N. Gomez, Lukasz Kaiser, and Illia Polosukhin. Attention is all you need. In I. Guyon, U. V. Luxburg, S. Bengio, H. Wallach, R. Fergus, S. Vishwanathan, and R. Garnett, editors, Advances in Neural I...

    2017

  31. [39]

    The Caltech-UCSD birds-200-2011 dataset

    Catherine Wah, Steve Branson, Peter Welinder, Pietro Perona, and Serge Be- longie. The Caltech-UCSD birds-200-2011 dataset. In Computation & Neural Systems Technical Report, number CNS-TR-2011-001, 2011

    2011

  32. [40]

    A deep semantic segmentation network with semantic and contextual re- finements

    Ziqiang Wang, Dong Liu, Li Yuan Wu, Sinuo Wang, Xiaoxiao Guo, and Lei Qi. A deep semantic segmentation network with semantic and contextual re- finements. arXiv preprint arXiv:2412.08671 , 2024

    2024 arXiv

  33. [41]

    Attention mechanisms in medical image segmentation: A survey

    Yutong Xie, Binjie Yang, Qing Guan, Jianpeng Zhang, Qiaolin Wu, and Yong Xia. Attention mechanisms in medical image segmentation: A survey. arXiv preprint arXiv:2305.17937, 2023

    2023 arXiv

  34. [42]

    A multi-channel and multi-spatial attention con- volutional neural network for prostate cancer isup grading

    Bo Yang and Zhi-tao Xiao. A multi-channel and multi-spatial attention con- volutional neural network for prostate cancer isup grading. Applied Sciences, 11(10):4321, 2021. 12 Hong Kong University of Science and Technology

    2021

  35. [43]

    From cnn to transformer: A review of medical image segmentation models

    Wenlong Yao, Jialun Bai, Weixin Liao, Yong Chen, Mingxia Liu, and Yakang Xie. From cnn to transformer: A review of medical image segmentation models. Journal of Imaging Informatics in Medicine , 37(4):1529–1547, Aug 2024

    2024

  36. [44]

    Toward understanding the effectiveness of attention mechanism

    Xiaofei Ye, Zijun He, Wee-Soon Heng, and Yu Li. Toward understanding the effectiveness of attention mechanism. AIP Advances, 13(3), 2023

    2023

  37. [45]

    Zhang, K

    H. Zhang, K. Dana, J. Shi, Z. Zhang, X. Wang, A. Tyagi, and A. Agrawal. Context encoding for semantic segmentation. In 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 7151–7160, Salt Lake City, UT, USA, 2018

    2018

  38. [46]

    Y. Zhou, X. Sun, Z. J. Zha, and W. Zeng. Context-reinforced semantic seg- mentation. In 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 4041–4050, Long Beach, CA, USA, 2019

    2019

  39. [47]

    Msp: Refine boundary segmentation via multiscale superpixel

    Jiacheng Zhu, Hanchi Huang, Bangjie Li, Yitong Liu, and Lisha Wang. Msp: Refine boundary segmentation via multiscale superpixel. arXiv preprint arXiv:2112.01746, 2021. Boris Kriuk Department of Computer Science and Engineering Hong Kong University of Science and Technology Hon...

    2021 arXiv

  40. [361]

    Springer, Cham, 2014

    2014

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.