pith. sign in

arxiv: 2605.21096 · v1 · pith:M6FQVZMHnew · submitted 2026-05-20 · 📡 eess.IV

Joint Alignment and Denoising for Event-Based Vision Sensors Using Regret-based Pareto Optimization

Shimpei Harada , Junya Hara , Hiroshi Higashi , Yuichi Tanaka This is my paper

Pith reviewed 2026-05-21 01:56 UTC · model grok-4.3

classification 📡 eess.IV
keywords event-based visionevent alignmentevent denoisingPareto optimizationcontrast mapregret strategymotion estimation
0
0 comments X

The pith

Joint regret-based Pareto optimization on contrast map variance aligns and denoises event-based vision data together.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper establishes that separate event alignment and denoising create a dilemma: noise biases alignment while unaligned events hinder denoising. It builds a contrast map that counts localized events per pixel, then poses alignment as maximizing the map's variance and denoising as minimizing that variance. These opposing goals are cast as a bi-objective Pareto problem solved with a regret strategy to locate a balanced operating point. Experiments on denoising and motion estimation tasks show the joint method outperforms pipelines that handle the two steps sequentially.

Core claim

The central claim is that a contrast map counting events per pixel can serve as the shared basis for both tasks, with variance maximization handling alignment and variance minimization handling denoising, and that regret-based Pareto optimization finds a practical solution to this bi-objective problem that improves downstream performance in event-based vision sensors.

What carries the argument

regret-based Pareto optimization on the variance of a contrast map that tallies events localized to each pixel

If this is right

  • Denoising improves because signal events are retained while noise events are suppressed through the shared variance objective.
  • Motion estimation accuracy rises because cleaner, better-aligned events feed into subsequent tracking or optical-flow algorithms.
  • The mutual bias between the two modules is reduced since neither step is performed in complete isolation from the other.
  • The same contrast-map formulation can be reused for related event-camera tasks that also depend on localized event density.

Where Pith is reading between the lines

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

  • The regret strategy could be replaced by other multi-objective solvers to test whether the performance lift is tied to the specific Pareto selection rule.
  • Extending the contrast map to include temporal decay or polarity might strengthen the variance proxy for high-speed scenes.

Load-bearing premise

That driving the contrast map variance in opposite directions for alignment and denoising produces a useful trade-off without introducing new biases or overlooking important event patterns.

What would settle it

A controlled test on synthetic or ground-truthed event streams where the joint method is compared against the strongest separate alignment-then-denoise or denoise-then-align pipelines and shows no gain or a loss in alignment accuracy or denoising metrics.

Figures

Figures reproduced from arXiv: 2605.21096 by Hiroshi Higashi, Junya Hara, Shimpei Harada, Yuichi Tanaka.

Figure 1
Figure 1. Figure 1: Top: Illustration of event drift and alignment of a [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: The trade-off between alignment and denoising. Align [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Scatter plots of denoised LEGO sequence in E–MLB dataset. Red points represent positive events; blue points represent [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Scatter plots of denoised hotel bar in DND21 dataset. Red points: detected signal events (TP). Yellow points: unde￾tected signal events (FP). Green points: detected noise events (TN). Black points: undetected noise events (FN). (a) BAF & CMax (b) EJA (c) Proposed [PITH_FULL_IMAGE:figures/full_fig_p007_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Scatter plots of aligned dynamic rotation in ECD dataset. Red points represent detected signal events ,while green points represent detected noise events. proposed method, we formulate these conflicting problems as a bi-objective Pareto optimization. This formulation provides a set of trade-off solutions between EA and ED. To control this trade-off and obtain a well-compromised solution, we use the regret … view at source ↗
read the original abstract

This paper proposes a joint alignment and denoising method for event-based vision sensors (EVSs). Existing signal processing methods for EVSs typically perform event alignment (EA) and event denoising (ED) as separate modules. However, this separation creates a dilemma: without ED, EA is biased by noise, whereas without EA, ED struggles to distinguish signal events from noise ones. To address this dilemma, we jointly optimize EA and ED by formulating a bi-objective Pareto optimization problem. Our formulation is built upon a contrast map that counts the number of events localized in each pixel. With a contrast map, we can formulate EA as maximizing its variance and ED as minimizing the variance. We cast these two conflicting problems as a Pareto optimization and use a regret strategy to obtain a solution. Experimental results on denoising and motion estimation demonstrate that our method achieves improvements against alternative ones.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 1 minor

Summary. The manuscript proposes a joint alignment and denoising method for event-based vision sensors. It formulates EA as maximizing the variance of an event-count contrast map and ED as minimizing the same variance, then solves the resulting bi-objective problem via a regret-based Pareto optimization strategy. The authors claim that this joint approach resolves the bias dilemma of separate EA/ED modules and yields improvements in denoising and motion estimation tasks over alternative methods.

Significance. If the central claim holds, the work would provide a concrete algorithmic route to handling the interdependence of alignment and denoising in event cameras, a recurring issue in neuromorphic vision pipelines. The regret-based Pareto formulation is a specific technical choice that could be reusable in other multi-objective event-processing settings. The significance is tempered by the need to confirm that the shared contrast-map variance objectives produce an unbiased trade-off rather than systematically suppressing signal events.

major comments (2)
  1. [Bi-objective formulation (contrast-map variance objectives)] The core premise that EA (maximize contrast-map variance) and ED (minimize the same variance) are usefully conflicting objectives is introduced when the contrast map is selected as the sole scalar field. Nothing in the formulation prevents the minimizer from discarding events that the maximizer would exploit for sharp alignment; this assumption is load-bearing for the joint-optimization claim and requires explicit evidence (e.g., ablation on event retention rates or orthogonality metrics between the two fronts) that the regret Pareto solution avoids systematic bias.
  2. [Abstract / Experimental results] The abstract asserts that experiments demonstrate improvements in denoising and motion estimation, yet supplies no quantitative metrics, datasets, error bars, or baseline details. Without these, the practical magnitude of the claimed gains cannot be evaluated and the central experimental support for the joint method remains unverifiable.
minor comments (1)
  1. [Method] Define the contrast map construction and the precise variance formula with an equation or pseudocode to support reproducibility.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive and detailed feedback on our manuscript. We address each major comment below with clarifications and indicate where revisions will be made to strengthen the presentation and evidence.

read point-by-point responses

  1. Referee: [Bi-objective formulation (contrast-map variance objectives)] The core premise that EA (maximize contrast-map variance) and ED (minimize the same variance) are usefully conflicting objectives is introduced when the contrast map is selected as the sole scalar field. Nothing in the formulation prevents the minimizer from discarding events that the maximizer would exploit for sharp alignment; this assumption is load-bearing for the joint-optimization claim and requires explicit evidence (e.g., ablation on event retention rates or orthogonality metrics between the two fronts) that the regret Pareto solution avoids systematic bias.

    Authors: We agree that explicit verification is needed to confirm the regret-based Pareto solution does not systematically suppress signal events. The regret formulation minimizes the maximum deviation from ideal single-objective optima, which is intended to produce a balanced front rather than allowing the minimizer to dominate. In the revision we will add an ablation study reporting event retention rates as a function of the regret parameter, together with a quantitative orthogonality measure (e.g., cosine similarity of the two objective gradients) between the alignment-maximizing and denoising-minimizing solutions. These additions will directly address the concern. revision: yes

  2. Referee: [Abstract / Experimental results] The abstract asserts that experiments demonstrate improvements in denoising and motion estimation, yet supplies no quantitative metrics, datasets, error bars, or baseline details. Without these, the practical magnitude of the claimed gains cannot be evaluated and the central experimental support for the joint method remains unverifiable.

    Authors: The abstract is intentionally concise; all quantitative results, including specific datasets (e.g., MVSEC and DVSNOISE20), metrics (PSNR, SSIM for denoising; angular and translational error for motion estimation), baseline comparisons, and error bars from multiple runs, appear in Section 4 of the manuscript. To improve accessibility we will revise the abstract to include one or two key numerical highlights and add explicit forward references to the experimental tables and figures. revision: partial

Circularity Check

0 steps flagged

No circularity: objectives defined directly from contrast map variance without reduction to inputs

full rationale

The paper defines its bi-objective problem explicitly by setting event alignment to maximize variance of the event-count contrast map and denoising to minimize the same variance, then applies regret-based Pareto optimization to the resulting trade-off. This is a direct modeling choice from the contrast map construction rather than any derivation that reduces by construction to fitted parameters, self-citations, or prior results. No load-bearing steps in the provided abstract or formulation rely on self-referential predictions or uniqueness theorems imported from the authors' own work. The experimental claims on denoising and motion estimation are presented as separate validation and do not feed back into the core optimization definition.

Axiom & Free-Parameter Ledger

1 free parameters · 1 axioms · 0 invented entities

Ledger is constructed from abstract only; full paper may introduce additional parameters or assumptions in the regret strategy or contrast-map definition.

free parameters (1)
  • regret parameter
    A tunable parameter in the regret strategy that balances the two objectives; its value is not specified in the abstract.
axioms (1)
  • domain assumption Variance of the contrast map serves as a valid proxy for both alignment quality and denoising effectiveness
    This premise is invoked when the paper formulates EA as maximizing variance and ED as minimizing variance of the same contrast map.

pith-pipeline@v0.9.0 · 5682 in / 1236 out tokens · 34119 ms · 2026-05-21T01:56:11.346632+00:00 · methodology

discussion (0)

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

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

40 extracted references · 40 canonical work pages · 1 internal anchor

  1. [1]

    Event- based vision: A survey,

    G. Gallego, T. Delbr ¨uck, G. Orchard, C. Bartolozzi, B. Taba, A. Censi, S. Leutenegger, A. J. Davison, J. Conradt, and K. Daniilidis, “Event- based vision: A survey,”IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 44, no. 1, pp. 154–180, 2020. TABLE III: RMSE↓on ECD dataset. Bold numbers indi- cate the best results while underlined n...

    work page 2020

  2. [2]

    Event Collapse in Contrast Maxi- mization Frameworks,

    S. Shiba, Y . Aoki, and G. Gallego, “Event Collapse in Contrast Maxi- mization Frameworks,”Sensors, vol. 22, no. 14, p. 5190, Jul. 2022

    work page 2022

  3. [3]

    A unifying contrast maximization framework for event cameras, with applications to motion, depth, and optical flow estimation,

    G. Gallego, H. Rebecq, and D. Scaramuzza, “A unifying contrast maximization framework for event cameras, with applications to motion, depth, and optical flow estimation,” inProceedings of the IEEE Confer- ence on Computer Vision and Pattern Recognition, 2018, pp. 3867–3876

    work page 2018

  4. [4]

    Frame-free dynamic digital vision,

    T. Delbruck, “Frame-free dynamic digital vision,” inProceedings of Intl. Symp. on Secure-Life Electronics, Advanced Electronics for Quality Life and Society, vol. 1. Citeseer, 2008, pp. 21–26

    work page 2008

  5. [5]

    Event density based denoising method for dynamic vision sensor,

    Y . Feng, H. Lv, H. Liu, Y . Zhang, Y . Xiao, and C. Han, “Event density based denoising method for dynamic vision sensor,”Applied Sciences, vol. 10, no. 6, p. 2024, 2020

    work page 2024

  6. [6]

    Neuromorphic Imaging with Density-based Spatiotemporal Denoising,

    P. Zhang, Z. Ge, L. Song, and E. Y . Lam, “Neuromorphic Imaging with Density-based Spatiotemporal Denoising,”IEEE Transactions on Computational Imaging, 2023

    work page 2023

  7. [7]

    Simultaneous Motion And Noise Estimation with Event Cameras,

    S. Shiba, Y . Aoki, and G. Gallego, “Simultaneous Motion And Noise Estimation with Event Cameras,” Aug. 2025

    work page 2025

  8. [8]

    Secrets of event-based optical flow, depth and ego-motion estimation by contrast maximiza- tion,

    S. Shiba, Y . Klose, Y . Aoki, and G. Gallego, “Secrets of event-based optical flow, depth and ego-motion estimation by contrast maximiza- tion,”IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 46, no. 12, pp. 7742–7759, 2024

    work page 2024

  9. [9]

    The theory of statistical decision,

    L. J. Savage, “The theory of statistical decision,”Journal of the American Statistical association, vol. 46, no. 253, pp. 55–67, 1951

    work page 1951

  10. [10]

    The use of reference objectives in multiobjective op- timization,

    A. P. Wierzbicki, “The use of reference objectives in multiobjective op- timization,” inMultiple criteria decision making theory and application: Proceedings of the third conference Hagen/K¨onigswinter, West Germany, August 20–24, 1979. Springer, 1980, pp. 468–486

    work page 1979

  11. [11]

    A 128×128 120 dB 15µ s latency asynchronous temporal contrast vision sensor,

    P. Lichtsteiner, C. Posch, and T. Delbruck, “A 128×128 120 dB 15µ s latency asynchronous temporal contrast vision sensor,”IEEE Journal of Solid-state Circuits, vol. 43, no. 2, pp. 566–576, 2008

    work page 2008

  12. [12]

    Entropy minimisation framework for event-based vision model estimation,

    U. M. Nunes and Y . Demiris, “Entropy minimisation framework for event-based vision model estimation,” inEuropean Conference on Computer Vision. Springer, 2020, pp. 161–176

    work page 2020

  13. [13]

    Live demonstration: Incremental motion estimation for event- based cameras by dispersion minimisation,

    ——, “Live demonstration: Incremental motion estimation for event- based cameras by dispersion minimisation,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2021, pp. 1322–1323

    work page 2021

  14. [14]

    Robust event-based vision model estimation by dispersion min- imisation,

    ——, “Robust event-based vision model estimation by dispersion min- imisation,”IEEE Transactions on Pattern Analysis and Machine Intel- ligence, vol. 44, no. 12, pp. 9561–9573, 2021

    work page 2021

  15. [15]

    A noise filtering algorithm for event-based asynchronous change detection image sensors on truenorth and its implementation on truenorth,

    V . Padala, A. Basu, and G. Orchard, “A noise filtering algorithm for event-based asynchronous change detection image sensors on truenorth and its implementation on truenorth,”Frontiers in Neuroscience, vol. 12, p. 118, 2018

    work page 2018

  16. [16]

    Design of a spatiotemporal correlation filter for event-based sensors,

    H. Liu, C. Brandli, C. Li, S.-C. Liu, and T. Delbruck, “Design of a spatiotemporal correlation filter for event-based sensors,” in2015 IEEE International Symposium on Circuits and Systems (ISCAS). IEEE, 2015, pp. 722–725

    work page 2015

  17. [17]

    O (N) O (N)-space spatiotemporal filter for reducing noise in neuromorphic vision sensors,

    A. Khodamoradi and R. Kastner, “O (N) O (N)-space spatiotemporal filter for reducing noise in neuromorphic vision sensors,”IEEE Trans- actions on Emerging Topics in Computing, vol. 9, no. 1, pp. 15–23, 2018

    work page 2018

  18. [18]

    Low cost and latency event camera background activity denoising,

    S. Guo and T. Delbruck, “Low cost and latency event camera background activity denoising,”IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 45, no. 1, pp. 785–795, 2022

    work page 2022

  19. [19]

    Probabilistic undirected graph based denoising method for dynamic vision sensor,

    J. Wu, C. Ma, L. Li, W. Dong, and G. Shi, “Probabilistic undirected graph based denoising method for dynamic vision sensor,”IEEE Trans- actions on Multimedia, vol. 23, pp. 1148–1159, 2020

    work page 2020

  20. [20]

    HashHeat: An O (C) complexity hashing-based filter for dynamic vision sensor,

    S. Guo, Z. Kang, L. Wang, S. Li, and W. Xu, “HashHeat: An O (C) complexity hashing-based filter for dynamic vision sensor,” in2020 25th Asia and South Pacific Design Automation Conference (ASP-DAC). IEEE, 2020, pp. 452–457. JOURNAL OF LATEX CLASS FILES, VOL. 18, NO. 9, SEPTEMBER 2020 8

    work page 2020

  21. [21]

    Effective target binarization method for linear timed address-event vision system,

    J. Xu, J. Zou, S. Yan, and Z. Gao, “Effective target binarization method for linear timed address-event vision system,”Optical Engineering, vol. 55, no. 6, pp. 063 103–063 103, 2016

    work page 2016

  22. [22]

    Event probability mask (epm) and event denoising convolutional neural network (edncnn) for neuromorphic cameras,

    R. Baldwin, M. Almatrafi, V . Asari, and K. Hirakawa, “Event probability mask (epm) and event denoising convolutional neural network (edncnn) for neuromorphic cameras,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2020, pp. 1701–1710

    work page 2020

  23. [23]

    EventZoom: Learn- ing to denoise and super resolve neuromorphic events,

    P. Duan, Z. W. Wang, X. Zhou, Y . Ma, and B. Shi, “EventZoom: Learn- ing to denoise and super resolve neuromorphic events,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2021, pp. 12 824–12 833

    work page 2021

  24. [24]

    DVS image noise removal using K-SVD method,

    X. Xie, J. Du, G. Shi, J. Yang, W. Liu, and W. Li, “DVS image noise removal using K-SVD method,” inNinth International Conference on Graphic and Image Processing (ICGIP 2017), vol. 10615. SPIE, 2018, pp. 1099–1107

    work page 2017

  25. [25]

    Event-based feature extraction using adaptive selection thresholds,

    S. Afshar, N. Ralph, Y . Xu, J. Tapson, A. van Schaik, and G. Cohen, “Event-based feature extraction using adaptive selection thresholds,” Sensors, vol. 20, no. 6, p. 1600, 2020

    work page 2020

  26. [26]

    Event stream super-resolution via spatiotemporal constraint learning,

    S. Li, Y . Feng, Y . Li, Y . Jiang, C. Zou, and Y . Gao, “Event stream super-resolution via spatiotemporal constraint learning,” inProceedings of the IEEE/CVF International Conference on Computer Vision, 2021, pp. 4480–4489

    work page 2021

  27. [27]

    A comprehensive survey of evolutionary-based multiobjective optimization techniques,

    C. A. Coello Coello, “A comprehensive survey of evolutionary-based multiobjective optimization techniques,”Knowledge and Information Systems, vol. 1, no. 3, pp. 269–308, 1999

    work page 1999

  28. [28]

    A tutorial on multiobjective opti- mization: Fundamentals and evolutionary methods,

    M. T. M. Emmerich and A. H. Deutz, “A tutorial on multiobjective opti- mization: Fundamentals and evolutionary methods,”Natural Computing, vol. 17, no. 3, pp. 585–609, Sep. 2018

    work page 2018

  29. [29]

    Multi-objective optimization using genetic algorithms: A tutorial,

    A. Konak, D. W. Coit, and A. E. Smith, “Multi-objective optimization using genetic algorithms: A tutorial,”Reliability engineering & system safety, vol. 91, no. 9, pp. 992–1007, 2006

    work page 2006

  30. [30]

    Many-Objective Evolutionary Algorithms: A Survey,

    B. Li, J. Li, K. Tang, and X. Yao, “Many-Objective Evolutionary Algorithms: A Survey,”ACM Computing Surveys, vol. 48, no. 1, pp. 1–35, Sep. 2015

    work page 2015

  31. [31]

    Survey of multi-objective optimization methods for engineering,

    R. T. Marler and J. S. Arora, “Survey of multi-objective optimization methods for engineering,”Structural and multidisciplinary optimization, vol. 26, no. 6, pp. 369–395, 2004

    work page 2004

  32. [32]

    A survey on search strategy of evolutionary multi-objective optimization algorithms,

    Z. Wang, Y . Pei, and J. Li, “A survey on search strategy of evolutionary multi-objective optimization algorithms,”Applied Sciences, vol. 13, no. 7, p. 4643, 2023

    work page 2023

  33. [33]

    A Tutorial on Evolution- ary Multiobjective Optimization,

    E. Zitzler, M. Laumanns, and S. Bleuler, “A Tutorial on Evolution- ary Multiobjective Optimization,” inMetaheuristics for Multiobjective Optimisation, G. Fandel, W. Trockel, X. Gandibleux, M. Sevaux, K. S ¨orensen, and V . T’kindt, Eds. Berlin, Heidelberg: Springer Berlin Heidelberg, 2004, vol. 535, pp. 3–37

    work page 2004

  34. [34]

    Adam: A Method for Stochastic Optimization

    D. P. Kingma, “Adam: A method for stochastic optimization,”arXiv preprint arXiv:1412.6980, 2014

    work page internal anchor Pith review Pith/arXiv arXiv 2014

  35. [35]

    E- MLB: Multilevel benchmark for event-based camera denoising,

    S. Ding, J. Chen, Y . Wang, Y . Kang, W. Song, J. Cheng, and Y . Cao, “E- MLB: Multilevel benchmark for event-based camera denoising,”IEEE Transactions on Multimedia, vol. 26, pp. 65–76, 2023

    work page 2023

  36. [36]

    The event-camera dataset and simulator: Event-based data for pose estimation, visual odometry, and slam,

    E. Mueggler, H. Rebecq, G. Gallego, T. Delbruck, and D. Scaramuzza, “The event-camera dataset and simulator: Event-based data for pose estimation, visual odometry, and slam,”The International Journal of Robotics Research, vol. 36, no. 2, pp. 142–149, 2017

    work page 2017

  37. [37]

    Front and back illuminated dynamic and active pixel vision sensors comparison,

    G. Taverni, D. P. Moeys, C. Li, C. Cavaco, V . Motsnyi, D. S. S. Bello, and T. Delbruck, “Front and back illuminated dynamic and active pixel vision sensors comparison,”IEEE Transactions on Circuits and Systems II: Express Briefs, vol. 65, no. 5, pp. 677–681, 2018

    work page 2018

  38. [38]

    Hots: A hierarchy of event-based time-surfaces for pattern recogni- tion,

    X. Lagorce, G. Orchard, F. Galluppi, B. E. Shi, and R. B. Benosman, “Hots: A hierarchy of event-based time-surfaces for pattern recogni- tion,”IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 39, no. 7, pp. 1346–1359, 2016

    work page 2016

  39. [39]

    EV- gait: Event-based robust gait recognition using dynamic vision sensors,

    Y . Wang, B. Du, Y . Shen, K. Wu, G. Zhao, J. Sun, and H. Wen, “EV- gait: Event-based robust gait recognition using dynamic vision sensors,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2019, pp. 6358–6367

    work page 2019

  40. [40]

    Ed- former: Transformer-based event denoising across varied noise levels,

    B. Jiang, B. Xiong, B. Qu, M. Salman Asif, Y . Zhou, and Z. Ma, “Ed- former: Transformer-based event denoising across varied noise levels,” inEuropean Conference on Computer Vision. Springer, 2024, pp. 200– 216

    work page 2024