Pith. sign in

REVIEW 4 major objections 6 minor 36 references

ViStruct: Simulating Expert-Like Reasoning Through Task Decomposition and Visual Attention Cues

T0 review · 4 major / 6 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read ViStruct claims that chart questions can be broken into structured subtasks, each tied to chart regions, producing expert-aligned reasoning.

desk verdict Useful prototype and a clear write-up, but the headline accuracy number is unsubstantiated and the expert-alignment claim outruns the evaluation. read the letter →

arxiv 2506.21762 v1 pith:TK7PT3S3 submitted 2025-06-26 cs.HC

Oliver Huang , Carolina Nobre This is my paper · ORCID
classification cs.HC
keywords taskdecompositionvisualattentionguidanceareasofinterestlargelanguagemodelsvision-languagevisualizationliteracychartcomprehensionexpertreasoning
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

ViStruct is an automated pipeline that claims to reproduce how expert chart readers reason: given a high-level question about a static chart, it breaks the question into a sequence of low-level subtasks and links each subtask to specific chart regions shown as visual attention cues. The paper reports 192 correct outputs in 225 trials across 45 tasks and 12 chart types, and expert reviewers rated the resulting guidance as clear, accurate, and aligned with their own reasoning. If the claim holds, implicit expert interpretation strategies can be externalized at scale, which matters for building visual-literacy tools and for studying how experts read charts.

What carries the argument

The central mechanism is the pairing of task decomposition with region-based visual attention. Decomposition is driven by a predefined taxonomy of ten low-level analytic task types; a breakdown–refine–verify prompting procedure turns a high-level question into atomic, executable subtasks. In parallel, computer-vision region detection identifies chart components such as axes, bars, and labels, a vision-language model assigns semantic labels to each region, and each subtask is tied to a set of areas of interest with spatial coordinates. The output is an editable workflow in which every reasoning step has a corresponding visual cue, such as a highlighted bar segment or a projected reference line to an axis.

What would settle it

Record expert readers' eye movements and think-aloud protocols on the same 45 tasks and compare their subtask sequences and attended regions against ViStruct's decompositions; if experts routinely reason in ways that cannot be segmented into the taxonomy's ten task types, or if their attention does not follow the system's areas of interest, the central claim of expert alignment would be contradicted. Alternatively, a controlled learning study showing that novices who follow ViStruct's guidance do not interpret charts better than unguided novices would undermine the usefulness of the generated reasoning sequences.

Watch

Extended reading notes

Core claim

The paper's central claim is that expert-like interpretation of data visualizations is reproducible by a fully automated pipeline. ViStruct first characterizes a chart structurally, then uses a prompted language model guided by a predefined taxonomy of ten low-level task types to decompose a high-level question into a validated sequence of subtasks, and then maps each subtask onto detected chart regions that are annotated as areas of interest. The pipeline is evaluated on 45 tasks spanning 12 chart types, with correct outputs in 192 of 225 trials (85.33%); a study with 20 trained visualization users produced mean ratings of 6.14/7 for guiding visual reasoning, 5.93/7 for accuracy of decompositions and areas of interest, and 5.97/7 for perceived expert-likeness. The authors present this as evidence that the system produces interpretable, expert-aligned reasoning sequences rather than as a direct instructional tool for novices.

Load-bearing premise

The load-bearing premise is that expert chart-reading reasoning reduces to a sequence of atomic subtasks drawn from a fixed ten-type taxonomy, with each subtask tied to one or more chart regions; the paper's own limitations state that more abstract tasks like correlation fit this pattern poorly.

Editorial extensions

If this is right

  • If ViStruct's central claim is correct, chart-reading guidance can be generated automatically for new charts and questions without hand-crafting each reasoning trace.
  • The same decomposition flow can make expert reasoning visible to novices and to researchers studying how charts are interpreted, since each step names both the operation and the region it applies to.
  • Because the system is model-agnostic, improvements in vision-language models should translate directly into more reliable region semantics and subtask grounding.
  • The reported failure pattern implies that concrete tasks such as value lookup and filtering are well served by region cues, while abstract tasks such as correlation need richer sources of evidence than highlighted areas alone.

Reading between the lines

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

  • A testable extension would be to measure whether novices who follow ViStruct's cues learn the underlying reasoning strategy and transfer it to a new chart, which the paper does not directly test.
  • If expert reasoning is not compositional in the way the taxonomy assumes, the pipeline's alignment with experts could be probed by comparing ViStruct's subtask sequences and attended regions against eye-tracking records of expert readers; such a comparison is not reported in the paper.
  • The authors' limitation that abstract tasks fit poorly suggests that a future design could condition guidance on task type, switching from region overlays to tooltip-style side panels when a task requires integrating multiple elements.
  • The editable workflow design implicitly treats reasoning as one valid path among several, which could be extended into an adaptive system that lets users steer decomposition interactively rather than accepting a fixed sequence.
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 / 6 minor

Summary. ViStruct is an automated pipeline that takes a static chart and a natural-language visualization question, produces a structured JSON chart description, decomposes the question into low-level subtasks using a ten-type taxonomy, detects chart regions with OpenCV and labels them via Gemini-2-Flash, and emits step-by-step visual attention cues. The paper reports a performance evaluation over 45 tasks and 225 trials with an 85.33% correctness figure, and a user study with 20 trained participants who rated usefulness, accuracy of decompositions and AOIs, and expert-likeness on Likert scales. The central claim is that the pipeline externalizes expert-like reasoning in an interpretable, expert-aligned form.

Significance. The paper has concrete strengths: the pipeline is model-agnostic and open-source, the breakdown-refine-verify prompting design is detailed, and the design goals G1-G5 are tied to prior work in visualization and cognitive science. If the evaluation were properly operationalized, ViStruct would be a useful prototype for visualization-literacy tools and for studying how decompositions can be grounded in chart regions. However, the current evidence does not establish the central claim: the headline accuracy figure is undefined, and the expert-alignment measure is self-report by the same population whose reasoning the system claims to simulate. These are correctable with a defined rubric, independent scoring, and a behavioral benchmark, so the contribution remains potentially valuable.

major comments (4)
  1. [Section 5.1] The paper states that "ViStruct produced correct outputs in 192 of the 225 trials (85.33%)" but never defines what counts as a correct output. The three evaluation criteria (G1 region identification, G2 value mapping, G3 decomposition coherence) are listed but not operationalized; no scoring rubric, no information about who performed the correctness judgments, and no inter-rater reliability are reported. Without this, the central quantitative claim cannot be verified or reproduced, and it is impossible to tell whether the 85.33% figure reflects pipeline quality or lenient judgment. Please provide a concrete rubric, an independent annotation procedure, and per-trial error analysis.
  2. [Section 5.2] Expert-alignment is measured by asking 20 trained visualization users to rate whether ViStruct's decompositions and AOIs align with their own reasoning. Because these participants are the target population whose reasoning the system claims to simulate, the measure is largely circular: a high rating shows that the system matches participants' stated preferences, not that it matches expert cognitive processes. The paper should add a behavioral comparison, for example, comparing ViStruct's step sequences and AOIs to eye-tracking data or to independently elicited expert protocols, and should report inter-rater agreement on the subjective ratings.
  3. [Section 6 (Limitations)] The limitations section acknowledges that AOI-based guidance is less effective for abstract tasks such as correlation and that OpenCV-based region-identification errors are a significant source of failure. These admissions directly qualify the 85.33% accuracy claim, yet Section 5.1 provides no breakdown of errors by task type or chart type. Please report per-task-type accuracy and an error taxonomy so that readers can see how abstract tasks and detection failures affect the headline figure.
  4. [Section 4.3] Decomposition is guided by a predefined taxonomy of ten low-level task types from Amar et al. [4]. The paper does not justify the assumption that expert visualization reasoning is fully expressible as sequences of these atomic task types, and the limitations section itself states that "AOIs alone are less effective for more abstract tasks." If expert reasoning is not compositional in this way, the expert-aligned claim fails even if the accuracy figure is correct. Please provide evidence for the coverage of the taxonomy on the 45 tasks, or analyze which tasks fall outside it.
minor comments (6)
  1. [Abstract and Conclusion] The phrase "confirming its ability to produce interpretable and expert-aligned reasoning sequences" is stronger than the evidence supports; the study measures perceptions, not objective alignment. Consider rewording to "suggest" or "indicate."
  2. [Section 4.1] There is a spacing typo in "In our initial experiments,We tested Gemini-2-Flash..." and the phrase "accessible here" appears without an actual link or citation to the repository.
  3. [Section 5.2] Participants were allowed to select any task from the question bank, which may bias the evaluation toward tasks they find comfortable; this selection mechanism should be reported and discussed as a potential source of bias.
  4. [Section 5.2] The Likert-scale results are reported as means and standard deviations without confidence intervals, effect sizes, or distribution information; given the large standard deviations (1.38-1.58), the paper should report more detailed descriptive statistics.
  5. [Figure 1] The pipeline overview in Figure 1 is dense; the labels (c1), (c2), (d), and (e) are difficult to map to the text in Section 4.1. Larger annotations or a numbered correspondence to the subsections would improve readability.
  6. [References] Several references are incomplete, such as those for Jamet et al. and Teo et al., which lack full bibliographic details; please standardize all entries.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the pipeline output and expert ratings are externally testable against VLAT/Mini-VLAT tasks and human judgment, with no fitted or self-referential step forcing the reported results.

full rationale

ViStruct is an empirical system paper rather than a derivation, so none of the classic circularity patterns apply. The headline accuracy figure (192/225 trials in Section 5.1) is a measured count on tasks from VLAT and Mini-VLAT; even though the paper does not specify a scoring rubric, nothing in the system construction forces that count. The 'expert-aligned' claim is tested with 20 trained users who rated usefulness, accuracy, and alignment with their own reasoning; these ratings could have been low and are therefore not tautological. The paper's stated limitations (e.g., 'AOIs alone are less effective for more abstract tasks') further show that outcomes are falsifiable. The only self-citations are background references to co-authored prior work on expert attention and productive friction, and these do not carry the performance or expert-approval results. No equation, fitted parameter, imported uniqueness theorem, or ansatz reduces the central claims to their inputs.

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

No new physical or theoretical entities are introduced. The central claim rests on modeling assumptions about expert reasoning and on the reliability of the LLM/VLM and OpenCV components, all listed as axioms.

assumptions (4)
  • domain assumption The ten-type low-level task taxonomy of Amar et al. [4] is sufficient to decompose any visualization question for the 12 chart types.
    Used as the sole decomposition guide in Section 4.3; if expert reasoning uses operations outside this taxonomy, the pipeline cannot represent them.
  • ad hoc to paper Following Areas of Interest (AOIs) in sequence approximates the reasoning process of experts.
    Design goals G3 and G5 assume sequential attention cues externalize expert reasoning, but the paper provides no eye-tracking or behavioral comparison, only subjective ratings.
  • domain assumption The vision-language model (Gemini-2-Flash) produces accurate JSON structural descriptions, region labels, and guidance steps.
    Every pipeline stage depends on the VLM's outputs; the overall 85.33% trial accuracy is the only aggregate check, and no per-stage reliability is reported.
  • ad hoc to paper Participants' self-reported alignment ratings measure actual expert-likeness of the system.
    Section 5.2 uses Likert ratings as validation of 'expert-aligned reasoning' without an independent criterion for expert behavior.

how reviews work

0 comments
Cite this review

Pith. "Pith review of ViStruct: Simulating Expert-Like Reasoning Through Task Decomposition and Visual Attention Cues." pith.science (2026). https://pith.science/paper/TK7PT3S3

@misc{pith2026250621762,
  author       = {Pith},
  title        = {Pith review of: ViStruct: Simulating Expert-Like Reasoning Through Task Decomposition and Visual Attention Cues},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/TK7PT3S3}},
  note         = {Machine review of arXiv:2506.21762}
}
read the original abstract

Data visualization tasks often require multi-step reasoning, and the interpretive strategies experts use, such as decomposing complex goals into smaller subtasks and selectively attending to key chart regions are rarely made explicit. ViStruct is an automated pipeline that simulates these expert behaviours by breaking high-level visual questions into structured analytic steps and highlighting semantically relevant chart areas. Leveraging large language and vision-language models, ViStruct identifies chart components, maps subtasks to spatial regions, and presents visual attention cues to externalize expert-like reasoning flows. While not designed for direct novice instruction, ViStruct provides a replicable model of expert interpretation that can inform the development of future visual literacy tools. We evaluate the system on 45 tasks across 12 chart types and validate its outputs with trained visualization users, confirming its ability to produce interpretable and expert-aligned reasoning sequences.

Figures

Figures reproduced from arXiv: 2506.21762 by the authors.

Figure 1
Figure 1. A pipeline overview of ViStruct. (a) User provides a visualization task description and static charts. (b) Behind the scenes, the system generates a detailed textual description of the chart and identifies its geometric elements. (c1) The high-level task is decomposed into a sequence of low-level component subtasks. (c2) An example shows how the original task is broken down into a step-by-step reasoning flow. (d) Th… view at source ↗
Figure 3
Figure 3. Workflow of region identification: The input chart is processed using OpenCV to identify distinct regions corresponding to chart segments. These regions are labeled by the system. VLM then identifies each annotated region with meaningful descriptions. 4.5 Region Identification Region identification is a key step in the ViStruct pipeline, providing the precise coordinate-level data needed for visual guidance and accu… view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

36 extracted references · 19 canonical work pages

  1. [4]

    R. Amar, J. Eagan, and J. Stasko. Low-level components of analytic activity in information visualization. In Proceedings of the Proceedings of the 2005 IEEE Symposium on Information Visualization, INFOVIS ’05, p. 15, 2005. doi: 10.1109/INFOVIS.2005.24 1, 2, 3

    doi:10.1109/infovis.2005.24 2005

  2. [1]

    Alicioglu and B

    G. Alicioglu and B. Sun. A survey of visual analytics for explainable artificial intelligence methods. Computers & Graphics, 102, 2022. doi: 10.1016/j.cag.2021.09.002 2

    doi:10.1016/j.cag.2021.09.002 2022

  3. [2]

    Alsaiari, J

    A. Alsaiari, J. Aurisano, and A. E. Johnson. Evaluating Strategies of Exploratory Visual Data Analysis in Multi Device Environments. In A. Kerren, C. Garth, and G. E. Marai, eds.,EuroVis 2020 - Short Papers. The Eurographics Association, 2020. doi: 10.2312/evs.20201054 2

    work page doi:10.2312/evs.20201054 2020

  4. [3]

    Alves, J

    A. Alves, J. Moura Pires, M. Y . Santos, A. Almeida, and A. León. Ai- assisted analytics–an automated approach to data visualization. InInter- national Conference on Conceptual Modeling, pp. 343–358. Springer,

    work page

  5. [5]

    Briakou, J

    E. Briakou, J. Luo, C. Cherry, and M. Freitag. Translating step-by-step: Decomposing the translation process for improved translation quality of long-form texts. In B. Haddow, T. Kocmi, P. Koehn, and C. Monz, eds., Proceedings of the Ninth Conference on Machine Translation, pp. 1301–1317, 2024. doi: 10.18653/v1/2024.wmt-1.123 1

    doi:10.18653/v1/2024.wmt-1.123 2024

  6. [6]

    T. T. Brunyé, T. Drew, K. F. Kerr, H. Shucard, D. L. Weaver, and J. G. Elmore. Eye tracking reveals expertise-related differences in the time-course of medical image inspection and diagnosis. 7. doi: 10. 1117/1.JMI.7.5.051203 2

    work page

  7. [7]

    Börner, A

    K. Börner, A. Bueckle, and M. Ginda. Data visualization literacy: Definitions, conceptual frameworks, exercises, and assessments. Pro- ceedings of the National Academy of Sciences , 116(6):1857–1864,

    work page

  8. [8]

    Y . Guo, D. Shi, M. Guo, Y . Wu, N. Cao, and Q. Chen. Talk2data: A natural language interface for exploratory visual analysis via question decomposition. ACM Trans. Interact. Intell. Syst., 14(2), article no. 8, 24 pages, Apr. 2024. doi: 10.1145/3643894 2

    doi:10.1145/3643894 2024

Show all 36 references
  1. [9]

    Y . Hou, B. Giledereli, Y . Tu, and M. Sachan. Do vision- language models really understand visual language? arXiv preprint arXiv:2410.00193, 2024. doi: 10.48550/arXiv.2410.00193 2

    2024 doi arXiv

  2. [10]

    Huang, Z

    Z. Huang, Z. Zhang, Z. Zha, Y . Lu, and B. Guo. Relationvlm: Mak- ing large vision-language models understand visual relations. CoRR, abs/2403.12801, 2024. doi: 10.48550/ARXIV.2403.12801 2

    2024 doi arXiv

  3. [11]

    Jamet, M

    E. Jamet, M. Gavota, and C. Quaireau. Attention guiding in multimedia learning. 18(2):135–145. Place: Netherlands Publisher: Elsevier Science. doi: 10.1016/j.learninstruc.2007.01.011 2

    2007 doi

  4. [12]

    Karer, I

    B. Karer, I. Scheler, H. Hagen, and H. Leitte. Conceptgraph: A formal model for interpretation and reasoning during visual analysis. Com- puter Graphics Forum, 2020. doi: 10.1111/cgf.13899 1

    2020 doi

  5. [13]

    Kazemitabaar, O

    M. Kazemitabaar, O. Huang, S. Suh, A. Z. Henley, and T. Grossman. Exploring the design space of cognitive engagement techniques with ai-generated code for enhanced learning. In Proceedings of the 30th International Conference on Intelligent User Interfaces, IUI ’25, 20 pages, ...

    2025

  6. [14]

    Kazemitabaar, J

    M. Kazemitabaar, J. Williams, I. Drosos, T. Grossman, A. Z. Henley, C. Negreanu, and A. Sarkar. Improving steering and verification in ai-assisted data analysis with interactive task decomposition. In Pro- ceedings of the 37th ACM Symposium on User Interface Software and Techn...

    2024

  7. [15]

    D. A. Keim. Information visualization and visual data mining. IEEE transactions on Visualization and Computer Graphics, 8, 2002. doi: 10 .1109/2945.981847 1

    2002

  8. [16]

    X. Li, Y . Sun, W. Cheng, Y . Zhu, and H. Chen. Chain-of-region: Visual language models need details for diagram analysis. In The Thirteenth International Conference on Learning Representations. 2

  9. [17]

    Moerth, Z

    E. Moerth, Z. Kostic, N. Gehlenborg, H. Pfister, J. Beyer, and C. Nobre. Beyond time and accuracy: Strategies in visual problem-solving. CHI ’25: Proceedings of the 2025 CHI Conference on Human Factors in Computing Systems, May 2025. doi: 10.1145/3706598.3714024 2

    2025

  10. [18]

    B. B. Morrison, L. E. Margulieux, and M. Guzdial. Subgoals, context, and worked examples in learning computing problem solving. In the 11th ACM International Computing Education Research Conference (ICER ’15), 2015. doi: 10.1145/2787622.2787733 1

    2015

  11. [19]

    Mukhopadhyay, A

    S. Mukhopadhyay, A. Qidwai, A. Garimella, P. Ramu, V . Gupta, and D. Roth. Unraveling the truth: Do vlms really understand charts? a deep dive into consistency and robustness. In Findings of the Association for Computational Linguistics: EMNLP 2024, pp. 16696–16717, 2024. doi:...

    2024 doi arXiv

  12. [20]

    A. A. Nelms and M. Segura-Totten. Expert–novice comparison reveals pedagogical implications for students’ analysis of primary literature. CBE–Life Sciences Education. doi: 10.1187/cbe.18-05-0077 1

    doi

  13. [21]

    Nobre, K

    C. Nobre, K. Zhu, E. Mörth, H. Pfister, and J. Beyer. Reading between the pixels: Investigating the barriers to visualization literacy. In Pro- ceedings of the 2024 CHI Conference on Human Factors in Computing Systems, CHI ’24, 2024. doi: 10.1145/3613904.3642760 2

    2024

  14. [22]

    Rezaie, M

    M. Rezaie, M. Tory, and S. Carpendale. Struggles and Strategies in Understanding Information Visualizations . IEEE Transactions on Visualization & Computer Graphics, 30(06):3035–3048, June 2024. doi: 10.1109/TVCG.2024.3388560 2

    2024

  15. [23]

    Schlieder, J

    A. Schlieder, J. Rummel, P. Albers, and F. Sadlo. Sequential visual cues from gaze patterns: Reasoning assistance for bar charts. InProceedings of the 2025 CHI Conference on Human Factors in Computing Systems, CHI ’25, 2025. doi: 10.1145/3706598.3713352 2

    2025

  16. [24]

    S. Suh, B. Min, S. Palani, and H. Xia. Sensecape: Enabling multilevel exploration and sensemaking with large language models. In Proceed- ings of the 36th annual ACM symposium on user interface software and technology, pp. 1–18, 2023. doi: 10.1145/3586183.3606756 1

    2023

  17. [25]

    Tankelevitch, V

    L. Tankelevitch, V . Kewenig, A. Simkute, A. E. Scott, A. Sarkar, A. Sellen, and S. Rintel. The metacognitive demands and opportunities of generative ai. In Proceedings of the 2024 CHI Conference on Human Factors in Computing Systems, CHI ’24, article no. 680, 24 pages, 2024. ...

    2024

  18. [26]

    T. W. Teo, Z. Q. Peh, T. W. Teo, and Z. Q. Peh. An exploratory study on eye-gaze patterns of experts and novices of science inference graph items. 3(3):205–229. doi: 10.3934/steme.2023013 1

    doi

  19. [27]

    J. G. Tullis, R. L. Goldstone, and A. J. Hanson. Scheduling scaffolding: The extent and arrangement of assistance during training impacts test performance. Journal of Motor Behavior, 47(5):442–452, 2015. 1

    2015

  20. [28]

    W. Wang, Y . Rao, R. Zhi, S. Marwan, G. Gao, and T. W. Price. Step tutor: Supporting students through step-by-step example-based feed- back. In Proceedings of the 2020 ACM Conference on Innovation and Technology in Computer Science Education, ITiCSE ’20, 2020. doi: 10 .1145/33...

    2020

  21. [29]

    J. Wei, X. Wang, D. Schuurmans, M. Bosma, F. Xia, E. Chi, Q. V . Le, D. Zhou, et al. Chain-of-thought prompting elicits reasoning in large language models. Advances in neural information processing systems, 35:24824–24837, 2022. doi: 10.5555/3600270.3602070 2

    2022

  22. [30]

    H. Xie, F. Wang, Y . Hao, J. Chen, J. An, Y . Wang, and H. Liu. The more total cognitive load is reduced by cues, the better retention and transfer of multimedia learning: A meta-analysis and two meta-regression analyses. 12(8):e0183884. doi: 10.1371/journal.pone.0183884 2

    doi

  23. [31]

    S. Yao, J. Zhao, D. Yu, N. Du, I. Shafran, K. Narasimhan, and Y . Cao. React: Synergizing reasoning and acting in language models. In International Conference on Learning Representations, 2023. doi: 10. 48550/arXiv.2210.03629 2

    2023 doi arXiv

  24. [32]

    J. Ye, A. Dash, W. Yin, and G. Wang. Beyond end-to-end VLMs: Leveraging intermediate text representations for superior flowchart understanding. In Proceedings of the Human Language Technologies (Volume 1), Apr. 2025. doi: 10.18653/v1/2025.naacl-long.180 2

    2025 doi

  25. [33]

    Zhang, W

    Y . Zhang, W. Zhang, Z. Zeng, K. Jiang, J. Li, W. Min, W. Luo, Q. Guan, J. Lin, and W. Yu. Mapreader: a framework for learning a visual lan- guage model for map analysis. International Journal of Geographical Information Science, 2025. doi: 10.1145/3557919.3565812 2

    2025

  26. [34]

    Y . Zhao, J. Wang, L. Xiang, X. Zhang, Z. Guo, C. Turkay, Y . Zhang, and S. Chen. Lightva: Lightweight visual analytics with llm agent- based task planning and execution. IEEE Transactions on Visualization and Computer Graphics, 2024. doi: 10.1109/TVCG.2024.3496112 2

    2024

  27. [2019]

    doi: 10.1073/pnas.1807180116 2

    doi

  28. [2024]

    doi: doi.org/10.1007/978-3-031-75599-6_24 2

    doi

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.