Pith. sign in

REVIEW 4 major objections 4 minor 59 references

Exploring Superpixel Segmentation Methods in the Context of Citizen Science and Deforestation Detection

T0 review · 4 major / 4 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read This paper claims that seven superpixel segmentation methods—RSS, ERGC, ETPS, CRS, LSC, SH, and GMMSP—perform quantitatively better than the SLIC baseline currently used in the ForestEyes citizen science project for deforestation detection.

desk verdict Useful applied comparison, but the headline ranking is only shown at one operating point and the gaps are small; it should be reviewed with a request for sensitivity analysis, not desk-rejected. read the letter →

arxiv 2411.17922 v3 pith:VTOMALDU submitted 2024-11-26 cs.CV

Hugo Resende , Isabela Borlido , Victor Sundermann , Eduardo B. Neto , Silvio Jamil F. Guimarães , Fabio Faria , Alvaro Luiz Fazenda This is my paper · ORCID
classification cs.CV
keywords superpixelsegmentationcitizenscienceForestEyesProjectdeforestationdetectionBrazilianLegalAmazonremotesensingSLICHomogeneityRate
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

ForestEyes is a citizen science project in which volunteers label image segments that may contain deforested areas in the Brazilian Amazon; the quality of those segments depends on the superpixel segmentation method. This paper compares 22 superpixel methods on nine Landsat-8 study areas, using classical quality metrics (boundary recall, undersegmentation error, color homogeneity, compactness, regularity, and SIRS) together with four new citizen-science-oriented metrics built on Homogeneity Rate. The central claim is that seven methods—RSS, ERGC, ETPS, CRS, LSC, SH, and GMMSP—outperform the SLIC baseline on the combined Final Score. If the claim holds, the project can switch segment generators to produce segments that volunteers label more accurately, which would improve downstream machine learning training.

What carries the argument

The central object is Homogeneity Rate ($HoR$), the percentage of pixels in a segment that belong to the majority class (forest or non-forest) under the official PRODES ground truth. Around $HoR$ the paper builds four citizen science metrics—Useful Segments (US, $HoR \geq 0.7$ and at least 70 pixels), Deforestation Segments (DS), perfect $HoR$ (PHoR), and pixel-level error (EP)—and combines average ranks over these and six classical metrics into a Final Score, where lower is better. This machinery turns raw segmentation outputs into an actionable winner list for volunteer-facing campaigns.

What would settle it

Re-run all 22 methods on the same nine study areas at several superpixel counts (for example 1000, 3000, 6000, and 12000) and recompute the Final Score; if the seven leading methods no longer consistently beat SLIC across budgets, the reported advantage is an artifact of the fixed 6000-superpixel setting. A simpler check is to compare the winning RSS and ERGC against SLIC using each method's own recommended number of segments.

Watch

Extended reading notes

Core claim

On the paper's own terms, the discovery is a ranking: when each method is scored by the mean of its ranks on six superpixel quality measures and four citizen science measures, seven of the 22 methods finish ahead of SLIC (RSS, ERGC, ETPS, CRS, LSC, SH, and GMMSP), with RSS and ERGC leading. The paper further proposes that a citizen science campaign treats a segment as useful when its Homogeneity Rate is at least 0.7 and it contains at least 70 pixels, and defines three companion metrics—deforestation share, perfect homogeneity share, and minority-class pixel error. The conclusion is that the ForestEyes project has an opportunity to improve its campaign quality by adopting one of these methods as a new baseline.

Load-bearing premise

Every method was forced to produce roughly 6000 superpixels, the maximum the deep method AINET could handle, so the comparison assumes that a single fixed superpixel budget is fair to all 22 methods.

Editorial extensions

If this is right

  • ForestEyes can adopt RSS, ERGC, ETPS, CRS, LSC, SH, or GMMSP as its segmentation baseline, since each scores higher than SLIC on the combined Final Score.
  • Campaign segments produced by the top methods should be more homogeneous, so volunteers face less ambiguous class mixtures and label more consistently.
  • The four citizen science metrics (US, DS, PHoR, and EP) give a reusable protocol for choosing segmentation methods in other volunteer-labeling projects.
  • The top five methods (RSS, ERGC, ETPS, CRS, and LSC) are slated for new citizen science campaigns to validate the ranking in real deployments.
  • Superpixel methods that favor boundary adherence, such as DISF, RSS, SH, and ISF, tend to sacrifice compactness, so a future winning method might improve compactness without losing delineation.

Reading between the lines

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

  • A natural extension the paper does not run is to test whether the citizen science ranking correlates with actual volunteer accuracy; if $HoR$-based metrics are the right proxy, the gap between SLIC and the top methods should show up in volunteer response agreement.
  • Because the fixed 6000-superpixel budget matched AINET's memory limit, methods that shine at very fine or very coarse scales may be unfairly ranked; per-method budget tuning could reshuffle the leaderboard.
  • The poor showing of deep superpixel networks (SSFCN, AINET, and SIN) likely reflects the domain gap between natural images and Landsat PCA bands; fine-tuning on remote sensing data might erase much of the gap.
  • If DS (deforestation segments among useful ones) is the tiebreaker, methods that produce many useful segments but few deforestation segments may be undervalued for monitoring recent deforestation, a separate objective from general segment quality.
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

4 major / 4 minor

Summary. The paper compares 22 superpixel segmentation methods on nine Landsat-8 images from the Brazilian Legal Amazon, in the context of the ForestEyes citizen-science project. The authors compute classical superpixel metrics (DS, BR, UE, SIRS, CO, Reg) and citizen-science-oriented metrics (DS, US, PHoR, EP), rank each method per metric, average the ranks into Score_SP and Score_CS, and average those into a Final Score (Table 3). The baseline is SLIC, the method currently used in ForestEyes. The central claim is that seven methods (RSS, ERGC, ETPS, CRS, LSC, SH, and GMMSP) outperform SLIC according to the Final Score.

Significance. If the result is robust, it offers a concrete, actionable improvement for the ForestEyes project and for similar citizen-science campaigns, and the paper's method coverage (22 methods spanning clustering, path-based, hierarchical, data-distribution, and deep approaches) is broader than many prior superpixel comparisons on remote sensing data. The proposed composite score is transparent, defined as an average of per-metric ranks rather than fitted to data, and the authors explicitly report the parameter choices used. The main risk is that the headline ranking is computed at one operating point (6000 superpixels) without sensitivity analysis or statistical tests, so the practical significance depends on the requested robustness checks.

major comments (4)
  1. [Section 4.1, Section 4.2.3, Table 3] Section 4.1 fixes the desired superpixel count at 6000 for all methods, chosen as the maximum feasible value for AINET, and this single operating point underlies the ranking in Table 3. Because the final score is the mean of rank means over only nine images and the margin between the seventh-ranked method (GMMSP, 10.483) and SLIC (10.733) is 0.25 rank points, the claim in Section 4.2.3 that seven methods outperform SLIC is not established as robust. Please provide a sensitivity analysis over a range of desired superpixel counts or otherwise justify that 6000 is the appropriate deployment count for all methods; at the least, report rankings at lower counts (e.g., 1000, 3000) and at the count actually used in past ForestEyes campaigns.
  2. [Section 4.2, Table 3] No uncertainty or significance measures accompany the final scores. Point estimates of rank means from nine images cannot support the comparative claim (e.g., GMMSP vs SLIC) without variance or paired tests. Report per-method standard deviations or bootstrap confidence intervals for Score_SP, Score_CS, and Final Score, and consider paired non-parametric tests (e.g., Wilcoxon signed-rank) comparing each method to SLIC on the underlying metric values.
  3. [Abstract, Section 1] The abstract and Section 1 state that 'most of the analyzed methods outperformed SLIC's performance,' but Table 3 shows only 7 of 22 methods with a Final Score below SLIC's 10.733, while 14 methods have higher scores. This overstatement should be corrected to 'seven methods' and applied consistently throughout the paper.
  4. [Section 4.1, Section 2.4] The evaluation fixes three task-specific parameters — the PCA input composition, the useful-segment HoR threshold of 0.7, and the minimum segment size of 70 pixels — based on earlier SLIC-oriented ForestEyes experiments [20, 21]. Applying choices that were tuned for SLIC to all 22 methods may bias the comparison toward methods that produce SLIC-like segment-size distributions. Please either justify these choices as task requirements independent of the segmentation method, or test the sensitivity of the final ranking to these thresholds.
minor comments (4)
  1. [Table 3, Section 5] The Final Score column shows LSC and SH tied at 9.900, but Section 5 names only LSC among the 'top five' methods; clarify the tie-breaking rule or list all six methods.
  2. [References] References [45] and [46] are the same Silvertown 2009 entry; remove the duplicate.
  3. [Table 3 caption/header] The Table 3 header appears to contain a typographical artifact ('Final ScoreDS'); the intended label should be 'Final Score DS' or 'Final Score'.
  4. [Section 4.2.2] The sentence 'CRS has the worst UE' is understandable, but for clarity consider stating that it has the highest undersegmentation error, since 'worst' could be misread as a low rank value.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the seven-method ranking is an empirical benchmark, and the self-cited thresholds/preprocessing do not force the outcome.

full rationale

The paper contains no derivation chain whose conclusion is equivalent to its premises. Section 4.1 defines the evaluation protocol: nine Landsat-8 study areas; all 22 methods run with parameters recommended by original authors; and the desired superpixel count fixed at 6000 for all methods due to AINET's memory limit. Section 4.2 computes classical metrics (BR, UE, SIRS, CO, Reg) and citizen-science metrics (US, DS, PHoR, EP), ranks each method per metric, and defines Score_SP, Score_CS, and Final Score as arithmetic means of ranks. These scores are working definitions, not fitted parameters, and no 'prediction' is generated from them other than reading Table 3. The claim in Section 4.2.3 that seven methods outperform SLIC is a direct consequence of the Final Score column, with RSS (8.733) lowest and SLIC (10.733) eleventh; this is an empirical comparison, not a self-fulfilling construction. The self-citations are limited to (i) the PCA band composition chosen because prior group work with SLIC showed improved performance, and (ii) the HoR>=0.7 and size>=70 thresholds for 'useful segments' taken from Dallaqua et al. [20]. Neither is fitted to the current nine images nor defined in terms of the reported ranking; they are fixed external criteria from earlier publications. Even if the thresholds or PCA choice favor certain method families, that is a validity/robustness matter, not circularity. The fixed 6000-superpixel operating point and the 0.25-rank gap between GMMSP and SLIC are legitimate sensitivity concerns but do not make the result equivalent to its inputs. No equation in the paper reduces to its own output, and no fitted value is renamed as a prediction. Hence no circular step is exhibited and the score is 0.

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

The central claim rests on several hand-chosen thresholds and representation choices (superpixel count, HoR threshold, segment size, PCA composition) that are imported from prior work or computational constraints. No new physical or algorithmic entities are introduced. The axioms are domain assumptions about metric validity, ground truth accuracy, parameter fairness, and dataset representativeness, none of which are empirically validated within the paper beyond the reported results.

free parameters (4)
  • Desired superpixel count = 6000
    Set to the maximum feasible value for the AINET method due to computational constraints; affects all methods equally but may not be optimal for each.
  • Homogeneity Rate threshold = 0.7
    Used to define Useful Segments; taken from prior ForestEyes work by Dallaqua et al. [20], not fitted here.
  • Minimum segment size = 70 pixels
    Segments smaller than 70 pixels are merged with their nearest neighbor; threshold from prior work [20].
  • PCA band composition = 3 principal components
    Selected based on a preliminary SLIC segmentation showing improved performance with PCA; applied to all methods.
assumptions (4)
  • domain assumption The citizen science metrics (US, DS, PHoR, EP) are valid proxies for volunteer interpretability and classification accuracy.
    The paper asserts these measures capture segment suitability for non-specialist volunteers, but no validation with actual volunteer responses is presented.
  • domain assumption PRODES ground truth is accurate and complete for the nine study areas.
    HoR values are computed against PRODES classifications; errors in ground truth would propagate to all metrics.
  • domain assumption Original author-recommended parameters provide a fair comparison across methods.
    The paper states parameters recommended by original authors are used to maintain fairness, but no sensitivity analysis is provided.
  • domain assumption The nine Landsat-8 images are representative of the Brazilian Legal Amazon for segmentation.
    All images are from the Xingu River Basin; results may not generalize to other regions or seasons.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Exploring Superpixel Segmentation Methods in the Context of Citizen Science and Deforestation Detection." pith.science (2026). https://pith.science/paper/VTOMALDU

@misc{pith2026241117922,
  author       = {Pith},
  title        = {Pith review of: Exploring Superpixel Segmentation Methods in the Context of Citizen Science and Deforestation Detection},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/VTOMALDU}},
  note         = {Machine review of arXiv:2411.17922}
}
read the original abstract

Tropical forests play an essential role in the planet's ecosystem, making the conservation of these biomes a worldwide priority. However, ongoing deforestation and degradation pose a significant threat to their existence, necessitating effective monitoring and the proposal of actions to mitigate the damage caused by these processes. In this regard, initiatives range from government and private sector monitoring programs to solutions based on citizen science campaigns, for example. Particularly in the context of citizen science campaigns, the segmentation of remote sensing images to identify deforested areas and subsequently submit them to analysis by non-specialized volunteers is necessary. Thus, segmentation using superpixel-based techniques proves to be a viable solution for this important task. Therefore, this paper presents an analysis of 22 superpixel-based segmentation methods applied to remote sensing images, aiming to identify which of them are more suitable for generating segments for citizen science campaigns. The results reveal that seven of the segmentation methods outperformed the baseline method (SLIC) currently employed in the ForestEyes citizen science project, indicating an opportunity for improvement in this important stage of campaign development.

Figures

Figures reproduced from arXiv: 2411.17922 by the authors.

Figure 1
Figure 1. Examples of segmentation results produced by SLIC, [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Examples of segments with different 𝐻𝑜𝑅 values. the derivation of various measures that enable analysis at both the segment and pixel levels. In this work, we proposed four measures to evaluate performance methods: (i) the percentage of so-called Useful Segments (US), which have 𝐻𝑜𝑅 ≥ 0.7 and a size of 70 pixels or more [20]; (ii) the percentage of deforestation segments among the useful segments (DS); (iii) the per… view at source ↗
Figure 3
Figure 3. The pipeline of the ForestEyes project. Highlighted is the target module of this work. [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: illustrates the study areas located in Pará, Brazil, high￾lighting key regions that require ongoing monitoring, such as in￾digenous lands and conservation units. The Xingu River Basin, rich in natural resources but vulnerable to illegal activities like mining, is given…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

59 extracted references · 48 canonical work pages

  1. [1]

    Radhakrishna Achanta, Appu Shaji, Kevin Smith, Aurelien Lucchi, Pascal Fua, and Sabine Süsstrunk. 2012. SLIC Superpixels Compared to State-of-the-Art Su- perpixel Methods. IEEE Transactions on Pattern Analysis and Machine Intelligence 34, 11 (2012), 2274–2282. https://doi.org/10.1109/TPAMI.2012.120

    doi:10.1109/tpami.2012.120 2012

  2. [2]

    Radhakrishna Achanta and Sabine Süsstrunk. 2017. Superpixels and Polygons Using Simple Non-iterative Clustering. In 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) . 4895–4904. https://doi.org/10.1109/CVPR. 2017.520

    doi:10.1109/cvpr 2017

  3. [3]

    David P Anderson, Jeff Cobb, Eric Korpela, Matt Lebofsky, and Dan Werthimer

    work page

  4. [4]

    Zhihua Ban, Jianguo Liu, and Li Cao. 2018. Superpixel Segmentation Using Gaussian Mixture Model. IEEE Transactions on Image Processing 27, 8 (2018), 4105–4117. https://doi.org/10.1109/TIP.2018.2836306

    work page arXiv 2018

  5. [5]

    Isabela Borlido Barcelos, Felipe De Castro Belém, Leonardo De Melo João, Ze- nilton K. G. Do Patrocínio, Alexandre Xavier Falcão, and Silvio Jamil Ferzoli Guimarães. 2024. A Comprehensive Review and New Taxonomy on Super- pixel Segmentation. ACM Comput. Surv. 56, 8, Article 200 (apr 2024), 39 pages. https://doi.org/10.1145/3652509

    work page doi:10.1145/3652509 2024

  6. [6]

    Isabela B Barcelos, Felipe De C Belém, Leonardo De M João, Alexandre X Falcão, and Guimarães Silvio JF. 2022. Improving color homogeneity mea- sure in superpixel segmentation assessment. In 2022 35th SIBGRAPI Confer- ence on Graphics, Patterns and Images (SIBGRAPI) , Vol. 1. IEEE, 79–84. https: //doi.org/10.1109/SIBGRAPI55357.2022.9991772

    arXiv 2022

  7. [7]

    Felipe Belém, Benjamin Perret, Jean Cousty, Silvio JF Guimarães, and Alexandre Falcão. 2022. Efficient Multiscale Object-based Superpixel Framework. arXiv preprint arXiv:2204.03533 (2022)

    work page Pith review arXiv 2022

  8. [8]

    Belém, Silvio Jamil F

    Felipe C. Belém, Silvio Jamil F. Guimarães, and Alexandre X. Falcão. 2020. Super- pixel Segmentation Using Dynamic and Iterative Spanning Forest. IEEE Signal Processing Letters 27 (2020), 1440–1444. https://doi.org/10.1109/LSP.2020.3015433

    arXiv 2020

Show all 59 references
  1. [9]

    Felipe C Belém, Benjamin Perret, Jean Cousty, Silvio JF Guimarães, and Alexan- dre X Falcão. 2021. Towards a Simple and Efficient Object-based Superpixel Delineation Framework. In 2021 34th SIBGRAPI Conference on Graphics, Patterns and Images (SIBGRAPI). IEEE, 346–353

    2021

  2. [10]

    Serge Bobbia et al. 2021. Iterative Boundaries implicit Identification for superpix- els Segmentation: a real-time approach. IEEE Access 9 (2021), 77250–77263

    2021

  3. [11]

    B. E. Boser, I. M. Guyon, and V. N. Vapnik. 1992. A training algorithm for optimal margin classifiers. In Workshop on Computational Learning Theory (Pittsburgh, Pennsylvania, United States) (COLT ’92). 144–152

    1992

  4. [12]

    Pierre Buyssens, Isabelle Gardin, and Su Ruan. 2014. Eikonal based region growing for superpixels generation: Application to semi-supervised real time organ segmentation in CT images. IRBM 35, 1 (2014), 20–26. https://doi.org/10. 1016/j.irbm.2013.12.007 Biomedical image segmen...

    2014

  5. [13]

    Dengfeng Chai. 2020. Rooted Spanning Superpixels. International Journal of Computer Vision 128, 12 (2020), 2962–2978

    2020

  6. [14]

    Jiansheng Chen, Zhengqin Li, and Bo Huang. 2017. Linear Spectral Clustering Superpixel. IEEE Transactions on Image Processing 26, 7 (2017), 3317–3330. https: //doi.org/10.1109/TIP.2017.2651389

    2017

  7. [15]

    Luis Pedro Coelho. 2012. Mahotas: Open source software for scriptable computer vision. arXiv preprint arXiv:1211.4907 (2012)

    2012 arXiv

  8. [16]

    Christian Conrad, Matthias Mertz, and Rudolf Mester. 2013. Contour-Relaxed Superpixels. In Energy Minimization Methods in Computer Vision and Pattern Recognition. Springer Berlin Heidelberg, Berlin, Heidelberg, 280–293

    2013

  9. [17]

    Seth Cooper, Firas Khatib, Adrien Treuille, Janos Barbero, Jeehyung Lee, Michael Beenen, Andrew Leaver-Fay, David Baker, Zoran Popović, et al. 2010. Predicting protein structures with a multiplayer online game. Nature 466, 7307 (2010), 756

    2010

  10. [18]

    Cover and P

    T. Cover and P. Hart. 1967. Nearest neighbor pattern classification. IEEE Trans- actions on Information Theory 13, 1 (1967), 21–27

    1967

  11. [19]

    Fernanda BJR Dallaqua, Álvaro L Fazenda, and Fabio A Faria. 2021. ForestEyes Project: Conception, enhancements, and challenges. Future Generation Computer Systems 124 (2021), 422–435

    2021

  12. [20]

    Fernanda Beatriz Jordan Rojas Dallaqua, Fabio Augusto Faria, and Álvaro Luiz Fazenda. 2020. Building Data Sets for Rainforest Deforestation Detection Through a Citizen Science Project. IEEE Geoscience and Remote Sensing Letters 19 (2020), 1–5

    2020

  13. [21]

    Fernanda Beatriz Jordan Rojas Dallaqua, Álvaro Luiz Fazenda, and Fabio Augusto Faria. 2019. ForestEyes project: Can citizen scientists help rainforests?. In 2019 15th International Conference on eScience (eScience) . IEEE, 18–27

    2019

  14. [22]

    Cesar Guerreiro Diniz, Arleson Antonio de Almeida Souza, Diogo Corrêa Santos, Mirian Correa Dias, Nelton Cavalcante Da Luz, Douglas Rafael Vidal De Moraes, Janaina Sant’Ana Maia, Alessandra Rodrigues Gomes, Igor da Silva Narvaes, Dalton M Valeriano, et al . 2015. DETER-B: The ...

    2015

  15. [23]

    Matthias Drusch et al. 2012. Sentinel-2: ESA’s optical high-resolution mission for GMES operational services. Remote sensing of Environment 120 (2012), 25–36

    2012

  16. [24]

    Fazenda and Fabio A

    Álvaro L. Fazenda and Fabio A. Faria. 2024. ForestEyes: Citizen Scientists and Machine Learning-Assisting Rainforest Conservation.Commun. ACM 67, 8 (2024), 95–96

    2024

  17. [25]

    Rémi Giraud, Vinh-Thong Ta, and Nicolas Papadakis. 2018. Robust superpixels using color and contour features along linear path. Computer Vision and Image Understanding 170 (2018), 1–13

    2018

  18. [26]

    Muki Haklay. 2013. Citizen science and volunteered geographic information: Overview and typology of participation. In Crowdsourcing geographic knowledge. Springer, 105–122

    2013

  19. [27]

    Mordechai Haklay, Daniel Dörler, Florian Heigl, Marina Manzoni, Susanne Hecker, Katrin Vohland, Katrin Vohland, Anne Land-Zandstra, L Ceccaroni, R Lemmens, et al. 2021. What is citizen science? The challenges of definition. The science of citizen science 13 (2021)

    2021

  20. [28]

    Robert M Haralick, Karthikeyan Shanmugam, et al. 1973. Textural features for image classification. IEEE Transactions on systems, man, and cybernetics 6 (1973), 610–621

    1973

  21. [29]

    Xuejing Kang, Lei Zhu, and Anlong Ming. 2020. Dynamic Random Walk for Superpixel Segmentation. IEEE Transactions on Image Processing 29 (2020), 3871–

    2020

  22. [30]

    David M Lapola, Patricia Pinho, Jos Barlow, Luiz EOC Aragão, Erika Berenguer, Rachel Carmenta, Hannah M Liddy, Hugo Seixas, Camila VJ Silva, Celso HL Silva-Junior, et al. 2023. The drivers and impacts of Amazon forest degradation. Science 379, 6630 (2023), eabp8622

    2023

  23. [31]

    Chris Lintott, Kevin Schawinski, Steven Bamford, Anže Slosar, Kate Land, Daniel Thomas, Edd Edmondson, Karen Masters, Robert C Nichol, M Jordan Raddick, et al. 2011. Galaxy Zoo 1: data release of morphological classifications for nearly 900 000 galaxies. Monthly Notices of the...

    2011

  24. [32]

    Ming-Yu Liu, Oncel Tuzel, Srikumar Ramalingam, and Rama Chellappa. 2011. Entropy rate superpixel segmentation. In CVPR 2011. 2097–2104. https://doi.org/ 10.1109/CVPR.2011.5995323

    2011

  25. [33]

    Lomax and S

    S. Lomax and S. Vadera. 2013. A Survey of Cost-sensitive Decision Tree Induction Algorithms. ACM Computing Surveys (CSUR)45, 2, Article 16 (mar 2013), 35 pages

    2013

  26. [34]

    D Lombraña González, A Harutyunyan, B Segal, I Zacharov, E McIntosh, PL Jones, M Giovannozzi, L Rivkin, MA Marquina, P Skands, et al . 2012. LHC@ Home: a volunteer computing system for massive numerical simulations of beam dynamics and high energy physics events. In Conf. Proc...

    2012

  27. [35]

    Eduardo FP Luz, Felipe RS Correa, Daniel L González, François Grey, and Fer- nando M Ramos. 2014. The forestwatchers: a citizen cyberscience project for deforestation monitoring in the tropics.Human Computation 1, 2 (2014), 137–145

    2014

  28. [36]

    Magdalena Main-Knorn et al. 2017. Sen2Cor for sentinel-2. In Image and signal processing for remote sensing XXIII , Vol. 10427. SPIE, 37–48

    2017

  29. [37]

    Nicolas Maisonneuve, Matthias Stevens, Maria E Niessen, and Luc Steels. 2009. NoiseTube: Measuring and mapping noise pollution with mobile phones. In Information technologies in environmental engineering . Springer, 215–228

    2009

  30. [38]

    N Justin Marshall, Diana A Kleine, and Angela J Dean. 2012. CoralWatch: educa- tion, monitoring, and sustainability through citizen science. Frontiers in Ecology and the Environment 10, 6 (2012), 332–334

    2012

  31. [39]

    David R Martin, Charless C Fowlkes, and Jitendra Malik. 2004. Learning to detect natural image boundaries using local brightness, color, and texture cues. IEEE Transactions on Pattern Analysis and Machine Intelligence 26, 5 (2004), 530–549. https://doi.org/10.1109/TPAMI.2004.1273918

    2004 arXiv

  32. [40]

    National Institute for Space Research (INPE). 2024. PRODES - Project for Moni- toring Deforestation in the Legal Amazon by Satellite. https://www.obt.inpe.br/ prodes/index.php. Accessed on: may, 2024

    2024

  33. [41]

    Peer Neubert and Peter Protzel. 2012. Superpixel benchmark and comparison. In Proc. Forum Bildverarbeitung, Vol. 6. 1–12

    2012

  34. [42]

    Global Forest Watch Pro, Forest W., and Forest Atlases. 2023. Global Forest Watch. Update (2023). Exploring Superpixel Segmentation Methods in the Context of Citizen Science and Deforestation Detection (Complete Version) Conference’17, July 2017, Washington, DC, USA

    2023

  35. [43]

    Eustaquio Reis and Rolando Guzman. 2023. An econometric model of Amazon deforestation. In The causes of tropical deforestation . Routledge, 172–191

    2023

  36. [44]

    Alexander Schick et al. 2012. Measuring and evaluating the compactness of super- pixels. In Proceedings of the 21st International Conference on Pattern Recognition (ICPR2012). IEEE, 930–934

    2012

  37. [46]

    Jonathan Silvertown. 2009. A new dawn for citizen science. Trends in ecology & evolution 24, 9 (2009), 467–471

    2009

  38. [47]

    Arfon M Smith, Stuart Lynn, and Chris J Lintott. 2013. An introduction to the zooniverse. In First AAAI conference on human computation and crowdsourcing

    2013

  39. [48]

    Patricia Smith, Oliver NF King, Avery Pennington, Win Tun, Mark Basham, Martin L Jones, Lucy M Collinson, Michele C Darrow, and Helen Spiers. 2023. Online citizen science with the Zooniverse for analysis of biological volumetric data. Histochemistry and Cell Biology 160, 3 (20...

    2023

  40. [49]

    Dave Stainforth, Jamie Kettleborough, Andrew P Martin, Andrew Simpson, Richard Gillis, Ali Akkas, Richard Gault, Mat Collins, David Gavaghan, and Myles Allen. 2002. Climateprediction. net: Design Principles for Publicresource Modeling Research.. In IASTED PDCS. Citeseer, 32–38

    2002

  41. [50]

    Subhashree Subudhi, Ram Narayan Patro, Pradyut Kumar Biswal, and Fabio Dell’Acqua. 2021. A Survey on Superpixel Segmentation as a Preprocessing Step in Hyperspectral Image Analysis. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing 14 (2021), 501...

    2021

  42. [51]

    Brian L Sullivan, Jocelyn L Aycrigg, Jessie H Barry, Rick E Bonney, Nicholas Bruns, Caren B Cooper, Theo Damoulas, André A Dhondt, Tom Dietterich, An- drew Farnsworth, et al. 2014. The eBird enterprise: an integrated approach to development and application of citizen science.B...

    2014

  43. [52]

    Michael Van den Bergh, Xavier Boix, Gemma Roig, and Luc Van Gool. 2015. Seeds: Superpixels extracted via energy-driven sampling. International Journal of Computer Vision 111, 3 (2015), 298–314. https://doi.org/10.1007/s11263-014- 0744-2

    2015 doi

  44. [53]

    John E Vargas-Muñoz, Ananda S Chowdhury, Eduardo B Alexandre, Felipe L Galvão, Paulo A Vechiatto Miranda, and Alexandre X Falcão. 2019. An iterative spanning forest framework for superpixel segmentation. IEEE Transactions on Image Processing 28, 7 (2019), 3477–3489

    2019

  45. [54]

    Yaxiong Wang, Yunchao Wei, Xueming Qian, Li Zhu, and Yi Yang. 2021. AINet: Association implantation for superpixel segmentation. In Proceedings of the IEEE/CVF International Conference on Computer Vision . 7078–7087

    2021

  46. [55]

    Xing Wei, Qingxiong Yang, Yihong Gong, Narendra Ahuja, and Ming-Hsuan Yang. 2018. Superpixel hierarchy. IEEE Transactions on Image Processing 27, 10 (2018), 4838–4849

    2018

  47. [56]

    Fengting Yang et al. 2020. Superpixel Segmentation With Fully Convolutional Networks. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

    2020

  48. [57]

    Jian Yao et al. 2015. Real-time coarse-to-fine topologically preserving segmenta- tion. In Proceedings of the IEEE conference on computer vision and pattern recogni- tion. 2947–2955

    2015

  49. [58]

    Qing Yuan, Songfeng Lu, Yan Huang, and Wuxin Sha. 2021. SIN: Superpixel Inter- polation Network. In Pacific Rim International Conference on Artificial Intelligence. Springer, 293–307

    2021

  50. [2002]

    SETI@ home: an experiment in public-resource computing. Commun. ACM 45, 11 (2002), 56–61

    2002

  51. [3884]

    https://doi.org/10.1109/TIP.2020.2967583

    2020

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.