Pith. sign in

REVIEW 3 major objections 4 minor 91 references

IRSAMap is a global dataset of 1.8 million vector-annotated land cover instances across 79 regions, built to move land cover mapping from pixel segmentation to object-based vector modeling.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

A new global remote sensing dataset with 1.8 million vector-annotated instances across 10 land cover classes, spanning 79 regions on six continents, for benchmarking vector-based land cover mapping.

T0 review reviewed 2026-08-05 challenge →

load-bearing objection Large-scale vector land cover dataset with real potential, but the abstract's accuracy guarantee is unproven and the 'first global' claim needs comparison against prior work. the 3 major comments →

arxiv 2508.16272 v1 pith:4U475PXS submitted 2025-08-22 cs.CV

IRSAMap:Towards Large-Scale, High-Resolution Land Cover Map Vectorization

Yu Meng , Ligao Deng , Zhihao Xi , Jiansheng Chen , Jingbo Chen , Anzhi Yue , Diyou Liu , Kai Li
show 4 more authors
Chenhao Wang Kaiyu Li Yupeng Deng Xian Sun
This is my paper
classification cs.CV
keywords land cover mappingvectorizationremote sensinginstance segmentationdataset benchmarkobject-based mappingtopologymulti-task learning
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

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 introduces IRSAMap, which it presents as the first global remote sensing dataset for large-scale, high-resolution, multi-feature land cover vector mapping. The dataset contains over 1.8 million instances of 10 typical objects—buildings, roads, rivers, and similar features—across 79 regions on six continents, totaling more than 1,000 km. The authors argue that existing datasets fall short because they carry few classes, small scale, and no spatial structure, while object-based vector mapping requires precise boundaries and topological consistency. IRSAMap is offered as a standardized benchmark that supports pixel-level classification, building outline extraction, road centerline extraction, and panoramic segmentation. If the dataset is as accurate as claimed, it gives the field a common testbed for the shift from pixel-based to object-based land cover modeling.

Core claim

The paper introduces IRSAMap as the first global remote sensing dataset for large-scale, high-resolution, multi-feature land cover vector mapping. It contains over 1.8 million instances of 10 typical objects—e.g., buildings, roads, rivers—across 79 regions on six continents, covering more than 1,000 km. The dataset is designed to have semantic and spatial accuracy through an intelligent annotation workflow that combines manual and AI-based methods. It supports multiple tasks: pixel-level classification, building outline extraction, road centerline extraction, and panoramic segmentation. The paper's claim is that this combination addresses the three limitations of existing datasets—limited cl

What carries the argument

The central object is IRSAMap itself: a vector annotation system in which objects are stored as polygons or lines with class labels, rather than as pixel grids. This vector representation is what preserves boundaries and topological structure. The intelligent annotation workflow combining manual and AI-based methods is the tool that makes it feasible to produce this structure at scale. The dataset's multi-task benchmark design carries the argument, because the same vector annotations can be used for pixel-level classification, building outlines, road centerlines, and panoramic segmentation, showing that vector structure is usable across tasks.

Load-bearing premise

The dataset's value depends on its annotations being semantically and spatially accurate, and the paper reports no verification metrics such as inter-annotator agreement or human-audit error rates.

What would settle it

Take a random sample of IRSAMap regions, have independent human experts redraw object polygons and labels, and measure boundary IoU and label agreement against the released annotations. If agreement falls well below typical inter-annotator levels for semantic segmentation, the accuracy claim is contradicted.

Watch this falsifier. Get emailed when new claim-graph text bears on it.

If this is right

  • A common benchmark for object-based land cover mapping becomes available, so methods can be compared on the same global, high-resolution vector data rather than on separate pixel-level datasets.
  • The same 1.8M-instance annotations can drive several tasks at once—pixel classification, building outline extraction, road centerline extraction, and panoramic segmentation—lowering the need for task-specific labeled data.
  • The six-continent coverage supports tests of geographic generalization and offers a foundation for automated global map updating.
  • Vector-format outputs move land cover mapping closer to GIS-ready results, which is directly relevant to digital twin construction.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Editorial inference: if the labels survive an independent audit, pretraining on IRSAMap could improve object-based mapping in regions outside the sampled 79, because the dataset's variety of object shapes and layouts may transfer better than region-specific datasets.
  • Editorial inference: the multi-task design implies a test the paper does not run—whether joint training on pixel labels, outlines, and centerlines improves each task relative to single-task baselines.
  • Editorial inference: a natural extension is temporal or multi-season versions of the same regions, which would turn the benchmark into a testbed for object-based change detection.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 4 minor

Summary. The paper introduces IRSAMap, described as the first global remote sensing dataset for large-scale, high-resolution, multi-feature land cover vector mapping. The abstract claims over 1.8 million instances of 10 object classes across 79 regions in six continents, an intelligent annotation workflow combining manual and AI-based methods, and multi-task adaptability. The central contribution is a publicly available benchmark intended to support the shift from pixel-level segmentation to object-based vector modeling. However, the abstract provides only qualitative assertions of annotation quality and no quantitative evaluation of label accuracy, annotation consistency, or dataset statistics.

Significance. If the stated scale and annotation quality hold, IRSAMap could be a valuable resource for object-based land cover mapping, enabling research on vectorization, boundary/topology modeling, and multi-task learning. The public release of the dataset is a concrete strength. However, the significance is conditional on the reliability of the ground-truth labels; without a quantitative accuracy assessment, the dataset cannot yet serve as a trustworthy benchmark. The work also has potential reproducibility value if the annotation workflow is documented in sufficient detail, which is not evidenced in the abstract.

major comments (3)
  1. [Abstract, Advantage 1] The central claim that the annotations have 'semantic and spatial accuracy' is load-bearing but entirely unsupported. No inter-annotator agreement, per-class accuracy/IoU, boundary error, or comparison against independent human-labeled references is reported. AI-assisted annotation is known to inherit model-specific biases (e.g., smooth boundaries, missed small objects, class confusion); 'consistency' of the workflow does not demonstrate correctness. The paper must include a quantitative validation protocol and results, otherwise the 1.8M instances may contain systematic errors that propagate into any model trained or evaluated on the dataset.
  2. [Abstract, Advantage 2] The 'intelligent annotation workflow combining manual and AI-based methods' is asserted but not specified. No details are given on the AI models used, the manual verification rate, conflict-resolution procedures, or quality-control thresholds. Without this information, the reproducibility of the workflow and the assessment of potential annotation biases are impossible. The manuscript should describe the workflow in detail and quantify the contribution of each stage to the final annotations.
  3. [Abstract, Advantage 3] The geographic coverage claim is imprecise: 'totaling over 1,000 km' lacks a correct area unit (likely km²), and no per-region or per-class statistics are provided. The claim of 'global coverage across 79 regions in six continents' is not substantiated by a breakdown of class distribution, region sizes, or sensor types. A dataset statistics table with number of instances, area, and class distribution per region is necessary to evaluate representativeness and potential geographic bias.
minor comments (4)
  1. [Abstract, Advantage 3] Typographical/units issue: 'totaling over 1,000 km' should likely be '1,000 km²' with the squared unit properly typeset.
  2. [Abstract, general] The phrase 'multi-feature' and 'multi-task adaptability' are not defined. Clarify what features are annotated (e.g., boundaries, centerlines, class labels) and which downstream tasks are supported.
  3. [Abstract, Advantage 1] '10 typical objects' is vague; provide the full class list in the abstract or, if space permits, a reference to a table in the full paper.
  4. [Related work (missing from abstract)] The claim of being 'the first global remote sensing dataset' for this task should be positioned against existing datasets such as DeepGlobe, SpaceNet, and others. A comparative table in the full paper would strengthen the novelty claim.

Circularity Check

0 steps flagged

No circularity: IRSAMap is a dataset announcement with no derivation chain, fitted predictions, or self-cited load-bearing results.

full rationale

The paper introduces a dataset and an annotation workflow. There is no mathematical derivation, no fitted parameter that is later called a prediction, and no model evaluation against the same data. The 'intelligent annotation workflow combining manual and AI-based methods' could in principle inherit model biases, but the abstract does not state that the same AI model is used for both annotation and evaluation, nor does it claim a quantitative prediction from a fitted model. The assertion of 'semantic and spatial accuracy' is presented as a property of the annotation process; its lack of quantitative validation is a correctness/transparency concern, not circularity. No self-citations or imported uniqueness theorems appear in the abstract. The dataset is publicly available, which makes external validation possible. Therefore, there is no significant circularity by the standards of this review.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The central contribution is a dataset, so there are no free parameters or invented entities. The load-bearing assumptions are the sufficiency of the 10-class taxonomy and, critically, the accuracy of the AI-assisted annotations that form the ground truth.

axioms (2)
  • domain assumption The 10 object classes (buildings, roads, rivers, etc.) constitute a sufficient taxonomy for land cover vector mapping.
    The abstract states that these 10 typical objects are annotated, implying they cover the relevant land cover categories. This is an assumption about the completeness of the class set.
  • domain assumption The AI-assisted annotation workflow produces labels accurate enough to serve as ground truth.
    The abstract claims 'ensuring semantic and spatial accuracy' but does not report verification metrics. The entire benchmark value rests on this assumption.

reviewed 2026-08-05 · how reviews work

0 comments
Cite this review

Pith. "Pith review of IRSAMap:Towards Large-Scale, High-Resolution Land Cover Map Vectorization." pith.science (2026). https://pith.science/paper/4U475PXS

@misc{pith2026250816272,
  author       = {Pith},
  title        = {Pith review of: IRSAMap:Towards Large-Scale, High-Resolution Land Cover Map Vectorization},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/4U475PXS}},
  note         = {Machine review of arXiv:2508.16272}
}
Share X Bluesky LinkedIn Reddit HN
read the original abstract

With the enhancement of remote sensing image resolution and the rapid advancement of deep learning, land cover mapping is transitioning from pixel-level segmentation to object-based vector modeling. This shift demands more from deep learning models, requiring precise object boundaries and topological consistency. However, existing datasets face three main challenges: limited class annotations, small data scale, and lack of spatial structural information. To overcome these issues, we introduce IRSAMap, the first global remote sensing dataset for large-scale, high-resolution, multi-feature land cover vector mapping. IRSAMap offers four key advantages: 1) a comprehensive vector annotation system with over 1.8 million instances of 10 typical objects (e.g., buildings, roads, rivers), ensuring semantic and spatial accuracy; 2) an intelligent annotation workflow combining manual and AI-based methods to improve efficiency and consistency; 3) global coverage across 79 regions in six continents, totaling over 1,000 km; and 4) multi-task adaptability for tasks like pixel-level classification, building outline extraction, road centerline extraction, and panoramic segmentation. IRSAMap provides a standardized benchmark for the shift from pixel-based to object-based approaches, advancing geographic feature automation and collaborative modeling. It is valuable for global geographic information updates and digital twin construction. The dataset is publicly available at https://github.com/ucas-dlg/IRSAMap

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

91 extracted references · 73 canonical work pages · 3 internal anchors

  1. [1]

    Jhawar, N

    M. Jhawar, N. Tyagi, and V. Dasgupta, ``Urban planning using remote sensing,'' International Journal of Innovative Research in Science, Engineering and Technology, vol. 1, no. 1, pp. 42--57, 2013

    2013

  2. [2]

    Das and D

    S. Das and D. P. Angadi, ``Land use land cover change detection and monitoring of urban growth using remote sensing and gis techniques: A micro-level study,'' GeoJournal, vol. 87, no. 3, pp. 2101--2123, 2022

    2022

  3. [3]

    Zhang, W

    Y. Zhang, W. Chen, B. Huang, Z. Zhang, J. Li, R. Gao, K. Wang, and C. Hu, ``An event logic graph for geographic environment observation planning in disaster chain monitoring,'' International Journal of Applied Earth Observation and Geoinformation, vol. 134, p. 104220, 2024

    2024

  4. [4]

    Benjdira, Y

    B. Benjdira, Y. Bazi, A. Koubaa, and K. Ouni, ``Unsupervised domain adaptation using generative adversarial networks for semantic segmentation of aerial images,'' Remote Sensing, vol. 11, no. 11, p. 1369, 2019

    2019

  5. [5]

    Z. Xu, W. Zhang, T. Zhang, and J. Li, ``Hrcnet: High-resolution context extraction network for semantic segmentation of remote sensing images,'' Remote Sensing, vol. 13, no. 1, p. 71, 2020

    2020

  6. [6]

    W. Liu, W. Zhang, X. Sun, Z. Guo, and K. Fu, ``Hecr-net: Height-embedding context reassembly network for semantic segmentation in aerial images,'' IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, vol. 14, pp. 9117--9131, 2021

    2021

  7. [7]

    Zhang, Z

    J. Zhang, Z. Zhou, G. Mai, M. Hu, Z. Guan, S. Li, and L. Mu, ``Text2seg: Remote sensing image semantic segmentation via text-guided visual foundation models,'' arXiv preprint arXiv:2304.10597, 2023

    arXiv 2023

  8. [8]

    B. Sui, Y. Cao, X. Bai, S. Zhang, and R. Wu, ``Bibed-seg: Block-in-block edge detection network for guiding semantic segmentation task of high-resolution remote sensing images,'' IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, vol. 16, pp. 1531--1549, 2023

    2023

  9. [9]

    X. Ma, R. Lian, Z. Wu, H. Guo, F. Yang, M. Ma, S. Wu, Z. Du, W. Zhang, and S. Song, ``Logcan++: Adaptive local-global class-aware network for semantic segmentation of remote sensing images,'' IEEE Transactions on Geoscience and Remote Sensing, 2025

    2025

  10. [10]

    Yaghmour, M

    A. Yaghmour, M. Crawford, and S. Prasad, ``A sensor agnostic domain generalization framework for leveraging geospatial foundation models: Enhancing semantic segmentation via synergistic pseudo-labeling and generative learning,'' in Proceedings of the Computer Vision and Pattern Recognition Conference, 2025, pp. 3047--3056

    2025

  11. [11]

    Maggiori, Y

    E. Maggiori, Y. Tarabalka, G. Charpiat, and P. Alliez, ``Convolutional neural networks for large-scale remote sensing image classification,'' IEEE Transactions on Geoscience and Remote Sensing, vol. 55, pp. 645--657, 2017

    2017

  12. [12]

    Sherrah, ``Fully convolutional networks for dense semantic labelling of high-resolution aerial imagery,'' arXiv preprint arXiv:1606.02585, 2016

    J. Sherrah, ``Fully convolutional networks for dense semantic labelling of high-resolution aerial imagery,'' arXiv preprint arXiv:1606.02585, 2016

    Pith/arXiv arXiv 2016

  13. [13]

    K. Yue, L. Yang, R. Li, W. Hu, F. Zhang, and W. Li, ``Treeunet: Adaptive tree convolutional neural networks for subdecimeter aerial image segmentation,'' ISPRS Journal of Photogrammetry and Remote Sensing, vol. 156, pp. 1--13, 2019

    2019

  14. [14]

    F. I. Diakogiannis, F. Waldner, P. Caccetta, and C. Wu, ``Resunet-a: A deep learning framework for semantic segmentation of remotely sensed data,'' ISPRS Journal of Photogrammetry and Remote Sensing, vol. 162, pp. 94--114, 2020

    work page 2020

  15. [15]

    Automatic Pixelwise Object Labeling for Aerial Imagery Using Stacked U-Nets

    A. Khalel and M. El-Saban, ``Automatic pixelwise object labeling for aerial imagery using stacked u-nets,'' arXiv preprint arXiv:1803.04953, 2018

    work page internal anchor Pith review Pith/arXiv arXiv 2018

  16. [16]

    R. Li, C. Duan, S. Zheng, C. Zhang, and P. M. Atkinson, ``Macu-net for semantic segmentation of fine-resolution remotely sensed images,'' IEEE Geoscience and Remote Sensing Letters, vol. 19, pp. 1--5, 2022

    work page 2022

  17. [17]

    Vaswani, N

    A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomez, . Kaiser, and I. Polosukhin, ``Attention is all you need,'' Advances in neural information processing systems, vol. 30, 2017

    2017

  18. [18]

    R. Xu, C. Wang, J. Zhang, S. Xu, W. Meng, and X. Zhang, ``Rssformer: Foreground saliency enhancement for remote sensing land-cover segmentation,'' IEEE Transactions on Image Processing, vol. 32, pp. 1052--1064, 2023

    work page 2023

  19. [19]

    W. Liu, N. Cui, L. Guo, S. Du, and W. Wang, ``Desformer: A dual-branch encoding strategy for semantic segmentation of very-high-resolution remote sensing images based on feature interaction and multi-scale context fusion,'' IEEE Transactions on Geoscience and Remote Sensing, 2024

    work page 2024

  20. [20]

    L. Wang, R. Li, C. Zhang, S. Fang, C. Duan, X. Meng, and P. M. Atkinson, ``Unetformer: A unet-like transformer for efficient semantic segmentation of remote sensing urban scene imagery,'' ISPRS Journal of Photogrammetry and Remote Sensing, vol. 190, pp. 196--214, 2022

    work page 2022

  21. [21]

    Batra, S

    A. Batra, S. Singh, G. Pang, S. Basu, C. Jawahar, and M. Paluri, ``Improved road connectivity by joint learning of orientation and segmentation,'' in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2019, pp. 10\,385--10\,393

    work page 2019

  22. [22]

    S. Wei, S. Ji, and M. Lu, ``Toward automatic building footprint delineation from aerial images using cnn and regularization,'' IEEE Transactions on Geoscience and Remote Sensing, vol. 58, no. 3, pp. 2178--2189, 2019

    work page 2019

  23. [23]

    Girard, D

    N. Girard, D. Smirnov, J. Solomon, and Y. Tarabalka, ``Polygonal building extraction by frame field learning,'' in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2021, pp. 5891--5900

    work page 2021

  24. [24]

    M \'a ttyus, W

    G. M \'a ttyus, W. Luo, and R. Urtasun, ``Deeproadmapper: Extracting road topology from aerial images,'' in Proceedings of the IEEE international conference on computer vision, 2017, pp. 3438--3446

    work page 2017

  25. [25]

    S. He, F. Bastani, S. Jagwani, M. Alizadeh, H. Balakrishnan, S. Chawla, M. M. Elshrif, S. Madden, and M. A. Sadeghi, ``Sat2graph: Road graph extraction through graph-tensor encoding,'' in Computer Vision--ECCV 2020: 16th European Conference, Glasgow, UK, August 23--28, 2020, Proceedings, Part XXIV 16. 1em plus 0.5em minus 0.4em Springer, 2020, pp. 51--67

    work page 2020

  26. [26]

    J. Li, J. He, W. Li, J. Chen, and J. Yu, ``Roadcorrector: A structure-aware road extraction method for road connectivity and topology correction,'' IEEE Transactions on Geoscience and Remote Sensing, 2024

    work page 2024

  27. [27]

    P. Yin, K. Li, X. Cao, J. Yao, L. Liu, X. Bai, F. Zhou, and D. Meng, ``Towards satellite image road graph extraction: A global-scale dataset and a novel method,'' in Proceedings of the Computer Vision and Pattern Recognition Conference, 2025, pp. 1527--1537

    work page 2025

  28. [28]

    Marcos, D

    D. Marcos, D. Tuia, B. Kellenberger, L. Zhang, M. Bai, R. Liao, and R. Urtasun, ``Learning deep structured active contours end-to-end,'' in Proceedings of the IEEE conference on computer vision and pattern recognition, 2018, pp. 8877--8885

    work page 2018

  29. [29]

    Cheng, R

    D. Cheng, R. Liao, S. Fidler, and R. Urtasun, ``Darnet: Deep active ray network for building segmentation,'' in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2019, pp. 7431--7439

    work page 2019

  30. [30]

    Hatamizadeh, D

    A. Hatamizadeh, D. Sengupta, and D. Terzopoulos, ``End-to-end trainable deep active contour models for automated image segmentation: Delineating buildings in aerial imagery,'' in European Conference on Computer Vision. 1em plus 0.5em minus 0.4em Springer, 2020, pp. 730--746

    work page 2020

  31. [31]

    J. Wang, L. Meng, W. Li, W. Yang, L. Yu, and G.-S. Xia, ``Learning to extract building footprints from off-nadir aerial images,'' IEEE transactions on pattern analysis and machine intelligence, vol. 45, no. 1, pp. 1294--1301, 2022

    work page 2022

  32. [32]

    Z. Xu, C. Xu, Z. Cui, X. Zheng, and J. Yang, ``Cvnet: Contour vibration network for building extraction,'' in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2022, pp. 1383--1391

    work page 2022

  33. [33]

    Huang, K

    X. Huang, K. Chen, Z. Wang, and X. Sun, ``Instance-aware contour learning for vectorized building extraction from remote sensing imagery,'' IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2024

    work page 2024

  34. [34]

    Zhang, S

    T. Zhang, S. Wei, Y. Zhou, M. Luo, W. Yu, and S. Ji, ``P2pformer: A primitive-to-polygon method for regular building contour extraction from remote sensing images,'' IEEE Transactions on Geoscience and Remote Sensing, 2024

    work page 2024

  35. [35]

    S. Wei, T. Zhang, S. Ji, M. Luo, and J. Gong, ``Buildmapper: A fully learnable framework for vectorized building contour extraction,'' ISPRS journal of photogrammetry and remote sensing, vol. 197, pp. 87--104, 2023

    work page 2023

  36. [36]

    Y. Hu, Z. Wang, Z. Huang, and Y. Liu, ``Polybuilding: Polygon transformer for building extraction,'' ISPRS Journal of Photogrammetry and Remote Sensing, vol. 199, pp. 15--27, 2023

    work page 2023

  37. [37]

    Z. Xu, Y. Liu, L. Gan, Y. Sun, X. Wu, M. Liu, and L. Wang, ``Rngdet: Road network graph detection by transformer in aerial images,'' IEEE Transactions on Geoscience and Remote Sensing, vol. 60, pp. 1--12, 2022

    work page 2022

  38. [38]

    Z. Xu, Y. Liu, Y. Sun, M. Liu, and L. Wang, ``Rngdet++: Road network graph detection by transformer with instance segmentation and multi-scale features enhancement,'' IEEE Robotics and Automation Letters, vol. 8, no. 5, pp. 2991--2998, 2023

    work page 2023

  39. [39]

    Hetang, H

    C. Hetang, H. Xue, C. Le, T. Yue, W. Wang, and Y. He, ``Segment anything model for road network graph extraction,'' in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2024, pp. 2556--2566

    work page 2024

  40. [40]

    Kirillov, E

    A. Kirillov, E. Mintun, N. Ravi, H. Mao, C. Rolland, L. Gustafson, T. Xiao, S. Whitehead, A. C. Berg, W.-Y. Lo et al., ``Segment anything,'' in Proceedings of the IEEE/CVF international conference on computer vision, 2023, pp. 4015--4026

    work page 2023

  41. [41]

    B. Yang, M. Zhang, Z. Zhang, Z. Zhang, and X. Hu, ``Topdig: Class-agnostic topological directional graph extraction from remote sensing images,'' in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 1265--1274

    work page 2023

  42. [42]

    Rottensteiner, G

    F. Rottensteiner, G. Sohn, M. Gerke, and J. D. Wegner, ``Isprs semantic labeling contest,'' ISPRS: Leopoldsh \"o he, Germany , vol. 1, no. 4, p. 4, 2014

    work page 2014

  43. [43]

    Volpi and D

    M. Volpi and D. Tuia, ``Dense semantic labeling of subdecimeter resolution images with convolutional neural networks,'' IEEE Transactions on Geoscience and Remote Sensing, vol. 55, no. 2, pp. 881--893, 2016

    work page 2016

  44. [44]

    Kampffmeyer, A.-B

    M. Kampffmeyer, A.-B. Salberg, and R. Jenssen, ``Semantic segmentation of small objects and modeling of uncertainty in urban remote sensing images using deep convolutional neural networks,'' in Proceedings of the IEEE conference on computer vision and pattern recognition workshops, 2016, pp. 1--9

    work page 2016

  45. [45]

    Marmanis, K

    D. Marmanis, K. Schindler, J. D. Wegner, S. Galliani, M. Datcu, and U. Stilla, ``Classification with an edge: Improving semantic image segmentation with boundary detection,'' ISPRS Journal of Photogrammetry and Remote Sensing, vol. 135, pp. 158--172, 2018

    work page 2018

  46. [46]

    Doi and A

    K. Doi and A. Iwasaki, ``The effect of focal loss in semantic segmentation of high resolution aerial image,'' in IGARSS 2018-2018 IEEE International Geoscience and Remote Sensing Symposium. 1em plus 0.5em minus 0.4em IEEE, 2018, pp. 6919--6922

    work page 2018

  47. [47]

    Zhang, I

    C. Zhang, I. Sargent, X. Pan, A. Gardiner, J. Hare, and P. M. Atkinson, ``Vprs-based regional decision fusion of cnn and mrf classifications for very fine resolution remotely sensed images,'' IEEE Transactions on Geoscience and Remote Sensing, vol. 56, no. 8, pp. 4507--4521, 2018

    work page 2018

  48. [48]

    G. Chen, X. Zhang, Q. Wang, F. Dai, Y. Gong, and K. Zhu, ``Symmetrical dense-shortcut deep fully convolutional networks for semantic segmentation of very-high-resolution remote sensing images,'' IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, vol. 11, no. 5, pp. 1633--1644, 2018

    work page 2018

  49. [49]

    Maggiori, Y

    E. Maggiori, Y. Tarabalka, G. Charpiat, and P. Alliez, ``Can semantic labeling methods generalize to any city? the inria aerial image labeling benchmark,'' in 2017 IEEE International geoscience and remote sensing symposium (IGARSS). 1em plus 0.5em minus 0.4em IEEE, 2017, pp. 3226--3229

    work page 2017

  50. [50]

    S. P. Mohanty, ``Crowdai dataset,'' 2018. [Online]. Available: https://www.crowdai.org/challenges/mapping-challenge/dataset_files

    work page 2018

  51. [51]

    Van Etten, D

    A. Van Etten, D. Lindenbaum, and T. M. Bacastow, ``Spacenet: A remote sensing dataset and challenge series,'' arXiv preprint arXiv:1807.01232, 2018

    Pith/arXiv arXiv 2018

  52. [52]

    Bastani, S

    F. Bastani, S. He, S. Abbar, M. Alizadeh, H. Balakrishnan, S. Chawla, S. Madden, and D. DeWitt, ``Roadtracer: Automatic extraction of road networks from aerial images,'' in Proceedings of the IEEE conference on computer vision and pattern recognition, 2018, pp. 4720--4728

    work page 2018

  53. [53]

    W. Jiao, C. Persello, and G. Vosselman, ``Polyr-cnn: R-cnn for end-to-end polygonal building outline extraction,'' ISPRS Journal of Photogrammetry and Remote Sensing, vol. 218, pp. 33--43, 2024

    work page 2024

  54. [54]

    Boguszewski, D

    A. Boguszewski, D. Batorski, N. Ziemba-Jankowska, T. Dziedzic, and A. Zambrzycka, ``Landcover. ai: Dataset for automatic mapping of buildings, woodlands, water and roads from aerial imagery,'' in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2021, pp. 1102--1110

    work page 2021

  55. [55]

    J. Wang, Z. Zheng, X. Lu, and Y. Zhong, ``Loveda: A remote sensing land-cover dataset for domain adaptive semantic segmentation,'' in Thirty-fifth Conference on Neural Information Processing Systems Datasets and Benchmarks Track (Round 2), 2021

    work page 2021

  56. [56]

    Tong, G.-S

    X.-Y. Tong, G.-S. Xia, Q. Lu, H. Shen, S. Li, S. You, and L. Zhang, ``Land-cover classification with high-resolution remote sensing images using transferable deep models,'' Remote Sensing of Environment, vol. 237, p. 111322, 2020

    work page 2020

  57. [57]

    Tong, G.-S

    X.-Y. Tong, G.-S. Xia, and X. X. Zhu, ``Enabling country-scale land cover mapping with meter-resolution satellite imagery,'' ISPRS Journal of Photogrammetry and Remote Sensing, vol. 196, pp. 178--196, 2023

    work page 2023

  58. [58]

    L. Wang, R. Li, D. Wang, C. Duan, T. Wang, and X. Meng, ``Transformer meets convolution: A bilateral awareness network for semantic segmentation of very fine resolution urban scene images,'' Remote Sensing, vol. 13, no. 16, p. 3065, 2021

    work page 2021

  59. [59]

    H. Chen, W. Yang, L. Liu, and G.-S. Xia, ``Coarse-to-fine semantic segmentation of satellite images,'' ISPRS Journal of Photogrammetry and Remote Sensing, vol. 217, pp. 1--17, 2024

    work page 2024

  60. [60]

    H. Guo, X. Su, C. Wu, B. Du, and L. Zhang, ``Building-road collaborative extraction from remote sensing images via cross-task and cross-scale interaction,'' IEEE Transactions on Geoscience and Remote Sensing, 2024

    work page 2024

  61. [61]

    Lu and Q

    X. Lu and Q. Weng, ``Multi-lora fine-tuned segment anything model for urban man-made object extraction,'' IEEE Transactions on Geoscience and Remote Sensing, 2024

    work page 2024

  62. [62]

    Demir, K

    I. Demir, K. Koperski, D. Lindenbaum, G. Pang, J. Huang, S. Basu, F. Hughes, D. Tuia, and R. Raskar, ``Deepglobe 2018: A challenge to parse the earth through satellite images,'' in 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), 2018, pp. 172--17\,209

    work page 2018

  63. [63]

    Weber and H

    E. Weber and H. Kan \'e , ``Building disaster damage assessment in satellite imagery with multi-temporal fusion,'' arXiv preprint arXiv:2004.05525, 2020

    Pith/arXiv arXiv 2004

  64. [64]

    Labs, ``Open cities ai challenge dataset,'' Accessed on: [Date Accessed], 2020, radiant MLHub

    G. Labs, ``Open cities ai challenge dataset,'' Accessed on: [Date Accessed], 2020, radiant MLHub. [Online]. Available: https://doi.org/10.34911/rdnt.f94cxb

    doi:10.34911/rdnt.f94cxb 2020

  65. [65]

    Q. Zhu, Y. Zhang, L. Wang, Y. Zhong, Q. Guan, X. Lu, L. Zhang, and D. Li, ``A global context-aware and batch-independent network for road extraction from vhr satellite imagery,'' ISPRS Journal of Photogrammetry and Remote Sensing, vol. 175, pp. 353--365, 2021

    work page 2021

  66. [66]

    J. Xia, N. Yokoya, B. Adriano, and C. Broni-Bediako, ``Openearthmap: A benchmark dataset for global high-resolution land cover mapping,'' in Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, 2023, pp. 6254--6264

    work page 2023

  67. [67]

    M. A. Friedl, D. Sulla-Menashe, B. Tan, A. Schneider, N. Ramankutty, A. Sibley, and X. Huang, ``Modis collection 5 global land cover: Algorithm refinements and characterization of new datasets,'' Remote sensing of Environment, vol. 114, no. 1, pp. 168--182, 2010

    work page 2010

  68. [68]

    Homer, C

    C. Homer, C. Huang, L. Yang, B. Wylie, and M. Coan, ``Development of a 2001 national land-cover database for the united states,'' Photogrammetric Engineering & Remote Sensing, vol. 70, no. 7, pp. 829--840, 2004

    work page 2001

  69. [69]

    J. Chen, J. Chen, A. Liao, X. Cao, L. Chen, X. Chen, C. He, G. Han, S. Peng, M. Lu et al., ``Global land cover mapping at 30 m resolution: A pok-based operational approach,'' ISPRS Journal of Photogrammetry and Remote Sensing, vol. 103, pp. 7--27, 2015

    work page 2015

  70. [70]

    LandCoverNet: A global benchmark land cover classification training dataset

    H. Alemohammad and K. Booth, ``Landcovernet: A global benchmark land cover classification training dataset,'' arXiv preprint arXiv:2012.03111, 2020

    work page internal anchor Pith review Pith/arXiv arXiv 2012

  71. [71]

    M. Luo, S. Ji, and S. Wei, ``A diverse large-scale building dataset and a novel plug-and-play domain generalization method for building extraction,'' IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, vol. 16, pp. 4122--4138, 2023

    work page 2023

  72. [72]

    Huang, K

    X. Huang, K. Chen, D. Tang, C. Liu, L. Ren, Z. Sun, R. H \"a nsch, M. Schmitt, X. Sun, H. Huang et al., ``Urban building classification (ubc) v2—a benchmark for global building detection and fine-grained classification from satellite imagery,'' IEEE Transactions on Geoscience and Remote Sensing, vol. 61, pp. 1--16, 2023

    work page 2023

  73. [73]

    H. Zhao, L. Fan, Y. Chen, H. Wang, X. Jin, Y. Zhang, G. Meng, Z.-X. Zhang et al., ``Opensatmap: A fine-grained high-resolution satellite dataset for large-scale map construction,'' Advances in Neural Information Processing Systems, vol. 37, pp. 59\,216--59\,235, 2024

    work page 2024

  74. [74]

    Haklay and P

    M. Haklay and P. Weber, ``Openstreetmap: User-generated street maps,'' IEEE Pervasive computing, vol. 7, no. 4, pp. 12--18, 2008

    work page 2008

  75. [75]

    L. Ding, H. Tang, and L. Bruzzone, ``Lanet: Local attention embedding to improve the semantic segmentation of remote sensing images,'' IEEE Transactions on Geoscience and Remote Sensing, vol. 59, no. 1, pp. 426--435, 2020

    work page 2020

  76. [76]

    K. Zhao, J. Kang, J. Jung, and G. Sohn, ``Building extraction from satellite images using mask r-cnn with building boundary regularization,'' in Proceedings of the IEEE conference on computer vision and pattern recognition workshops, 2018, pp. 247--251

    work page 2018

  77. [77]

    T. Y. Zhang and C. Y. Suen, ``A fast parallel algorithm for thinning digital patterns,'' Communications of the ACM, vol. 27, no. 3, pp. 236--239, 1984

    work page 1984

  78. [78]

    Cheng, Y

    G. Cheng, Y. Wang, S. Xu, H. Wang, S. Xiang, and C. Pan, ``Automatic road detection and centerline extraction via cascaded end-to-end convolutional neural network,'' IEEE Transactions on Geoscience and Remote Sensing, vol. 55, no. 6, pp. 3322--3337, 2017

    work page 2017

  79. [79]

    B. Xu, J. Xu, N. Xue, and G.-S. Xia, ``Hisup: Accurate polygonal mapping of buildings in satellite imagery with hierarchical supervision,'' ISPRS Journal of Photogrammetry and Remote Sensing, vol. 198, pp. 284--296, 2023

    work page 2023

  80. [80]

    C. Wang, J. Chen, Y. Meng, Y. Deng, K. Li, and Y. Kong, ``Sampolybuild: Adapting the segment anything model for polygonal building extraction,'' ISPRS Journal of Photogrammetry and Remote Sensing, vol. 218, pp. 707--720, 2024

    work page 2024

Showing first 80 references.

This paper was first reviewed by deepseek-v4-flash on August 5, 2026.