Pith. sign in

REVIEW 3 major objections 5 minor 32 references

Escher Tile Deformation via Closed-Form Solution

T0 review · 3 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

Pith's one-line read Deforming an Escher tile while preserving its tessellation reduces to summing a geometric-series falloff over every translated copy of a user's point handle, giving a closed-form displacement field that works for all 17 wallpaper groups.

desk verdict Clever closed-form deformation for Escher tiles, but the tileability guarantee is conditional on injectivity and is not proven. read the letter →

arxiv 2506.23388 v1 pith:LCIOAXKO submitted 2025-06-29 cs.GR cs.CGcs.MSmath.MG

classification cs.GRcs.CGcs.MSmath.MG MSC 52C2068U05
keywords Eschertileswallpapergroupsclosed-formsolutionperiodicdisplacementfieldpointhandledeformationBravaislatticegeometricseriestessellation
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

This paper sets out to show that interactive deformation of periodic Escher tiles can be done exactly, in real time, and without breaking the tessellation. The user drags a point handle, and the algorithm sums that handle's influence over all infinitely repeated copies of the tile to produce a periodic displacement field. The key move is choosing a falloff whose infinite lattice sum is a geometric series, which yields closed-form weights for the four lattice types that cover all 17 wallpaper groups. If correct, artists can reshape a single tile and immediately see the whole pattern adjust so that every copy still locks together with no gaps or overlaps, with boundary and interior deforming together.

What carries the argument

The central object is the guidance field: a single fundamental vector, generated by the wallpaper group's cone points and mirror lines from the user's handle displacement, repeated by pure translation at every vertex of the governing Bravais lattice. Around each handle the paper places an exponential falloff whose exponents are sums of absolute values, chosen so that the infinite sum over all lattice copies collapses into a geometric series. Splitting the sum by quadrants and evaluating once per unit cell produces the closed-form weights of Equations (5), (11), and (14). The displacement field is then a finite linear combination of fundamental vectors with those weights, making the deformation representation-independent, real-time, and periodic by construction.

What would settle it

Take a tile of wallpaper group 2222 on a parallelogram lattice, set the falloff parameter so the deformation reaches across several cells, drag the handle by more than one cell width, and check whether any two sample points satisfy $x + u(x) = y + u(y)$; equivalently, sample the Jacobian determinant of $x \mapsto x + u(x)$ in one cell and look for a sign change, which would expose a fold and contradict the no-gaps/no-overlaps claim.

Watch

Extended reading notes

Core claim

The central claim is that deforming a tessellation while preserving its tileability can be reduced to a closed-form periodic displacement field. From one user handle displacement, the wallpaper group's cone points and mirror lines generate a small set of fundamental vectors, and pure translation then repeats each fundamental vector at every vertex of the governing Bravais lattice. The influence of all those copies at any point is a weight that can be evaluated analytically because the paper's exponential falloff makes the infinite double sum a geometric series. Equations (5), (11), and (14) give the closed-form weights for the tetragonal, hexagonal, and monoclinic lattices, with the orthorhombic case reusing the tetragonal formula, so the final displacement at any point is a finite weighted sum of fundamental vectors. Because the field is periodic by construction, the deformed pattern remains a candidate tessellation, and the method applies uniformly to images, meshes, and volumetric cells.

Load-bearing premise

The load-bearing premise is that the map $x \mapsto x + u(x)$ that sends every point to its deformed position is injective, so two distinct points never land on the same spot; the paper assumes this rather than proving it and lists enforcing it as the first item of future work.

Editorial extensions

If this is right

  • An artist can edit a tile's boundary and interior texture in one gesture, seeing the full appearance change immediately instead of editing outlines and filling textures separately.
  • Because the weight for each fundamental vector is computed once in a unit cell, the cost is independent of how many copies of the tile are visible on screen.
  • The adaptive falloff parameter gives users a continuous locality dial, from a sharp local edit to a global wave affecting the entire tessellation.
  • The same closed-form construction extends from 2D images and meshes to 3D parallelepiped cells, as demonstrated by sculpting one packable animal shape into another.
  • Symmetry transitions between wallpaper groups let a deformation lower a pattern's symmetry level, enabling artistic exploration across related tilings.

Reading between the lines

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

  • Because the geometric-series falloff is not divergence-free, very large handle drags can fold the map $x \mapsto x + u(x)$; swapping the falloff for a divergence-free fundamental solution of the elastic wave equation would make the deformation provably injective.
  • The same recipe of decoupling the point-group orbit from the translation lattice should transfer to tessellations of the sphere or hyperbolic plane, where the translation lattice is replaced by a non-Euclidean group and the geometric series by a group-theoretic sum.
  • The weight functions resemble lattice Green's functions, so sharper or Gaussian falloffs may be expressible with theta functions, widening the locality-versus-closed-form tradeoff without sacrificing speed.
  • A direct experimental check of injectivity—sampling the Jacobian determinant of $x \mapsto x + u(x)$ across a unit cell and monitoring for sign changes—would tell artists exactly how far a handle can be dragged before the no-gaps/no-overlaps promise breaks.
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

3 major / 5 minor

Summary. This paper proposes a real-time interactive deformation method for Escher-style periodic tilings. A user drags a point handle; the method generates fundamental displacement vectors from the tile's symmetry group and computes a periodic displacement field by summing a geometric-series falloff kernel over all lattice copies of the handle. Closed-form expressions are derived for square, rectangular, hexagonal, and parallelogram lattices (Eqs. 5, 11, 14), enabling deformation of image, mask, and mesh representations. The central claim is that this deformation simultaneously adjusts boundaries and interiors while preserving tileability — no gaps or overlaps — for all 17 wallpaper groups.

Significance. The closed-form evaluation of the infinite lattice sums is elegant, and the real-time rendering claim is plausible; the comparison against truncated numerical summation in Fig. 5 at least checks the algebra of the series. If the tileability guarantee were established, this would be a practical contribution to interactive pattern design, especially because it handles textured interiors rather than only boundaries. However, the guarantee as stated is not established: the deformation map is nowhere shown to be injective, Section 6 explicitly defers injectivity to future work, and the presented validation cannot detect gaps or overlaps. The current significance is therefore as a well-motivated deformation system with attractive closed-form weights, but with a load-bearing correctness claim left open.

major comments (3)
  1. [Sections 4 and 6] The paper's headline claim is that the deformation preserves tileability 'without introducing gaps or overlaps,' but Section 6 lists 'ensure injectivity of the map under extreme deformations' as the first future-work item. Since the deformation map is implicitly x -> x + u(x), tileability requires this map to be injective (equivalently, local injectivity with det(I + nabla u) > 0 under the periodic setting). The displacement field is a linear combination of handle drag vectors (Eq. 1), so its gradient scales with the drag magnitude; for any sigma whose kernel has a nonzero gradient, sufficiently large user drags will make det(I + nabla u) negative and fold the pattern. The paper gives no bound on handle displacement or sigma within which the headline guarantee holds, so Eqs. (5), (11), and (14) currently define a displacement field, not a proven tileability-preserving deformation.
  2. [Figure 5 and Section 4] The validation in Fig. 5 compares the closed-form expressions with a truncated numerical summation of the same infinite series. This verifies only the algebraic evaluation of the sum; it does not test whether the resulting map is injective, whether tile boundaries remain matched after applying the displacement, or whether interior overlaps occur. No quantitative metric for gaps or overlaps is reported, so this comparison cannot support the 'without gaps or overlaps' claim.
  3. [Sections 4.1-4.4] The paper claims support for all 17 wallpaper groups, but the exposition provides closed-form weights for four lattice geometries and illustrates only a few groups. No table or algorithm specifies, for each of the 17 groups, the number and directions of the fundamental vectors or the linear transformations applied to u0 for rotations, reflections, and glide reflections, nor is there a proof that the resulting field satisfies u(g(x)) = g(u(x)) for every symmetry operation g of the group. Without this per-group specification, the 'all 17 wallpaper groups' claim is not substantiated, and the symmetry-transition examples in Section 5 remain purely visual.
minor comments (5)
  1. [Section 4] The deformation map is never written explicitly; near the definition of u, the paper should state x' = x + u(x) so that the reader knows the final transformation being applied.
  2. [Eq. (4)] The local-coordinate formula uses unexplained double bars and a modulo operation; the exact definition of the map from global to local coordinates should be clarified.
  3. [Eqs. (2), (5), (14)] The admissible range of sigma is not discussed. The closed-form denominators contain (2^sigma - 1)^2, which vanishes as sigma -> 0, so the paper should state the valid range of sigma and the behavior of the solution as sigma approaches singular values.
  4. [Section 4.4] There are typos: 'monoiclinic' should be 'monoclinic,' 'feaaturing' should be 'featuring,' and 'Bravias lattice' should be 'Bravais lattice.'
  5. [Section 5] The 'Symmetry Transitions' examples are presented as demonstrating symmetry downgrades, but the subgroup relations are not verified or even enumerated; the text should either state the relevant subgroup inclusions or label the results as informal visual observations.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the closed-form weights are analytic evaluations of explicitly defined geometric-series falloffs, and the validation is a self-consistency check rather than a fitted prediction.

full rationale

The derivation chain is self-contained. The paper defines a point-handle displacement as a weighted sum of fundamental vectors (Eq. 1), selects an explicit exponential falloff (Eq. 2) so that the infinite lattice sum is a geometric series, and then evaluates that sum analytically (Eqs. 5, 11, 14). The closed-form expressions are not fitted to data and are not imported from prior work; they are direct evaluations of the paper's own definitions. The comparative validation in Figure 5 checks the analytic expression against a truncated numerical summation of the same series, which is a consistency check of the algebra rather than an independent prediction, and it is not presented as evidence that tileability is preserved. The paper's central claim about avoiding gaps and overlaps depends on the undeclared assumption that x + u(x) is injective; Section 6 explicitly lists 'ensure injectivity of the map under extreme deformations' as a limitation and future work. That is a soundness gap, not circular reasoning: the displacement field is still well-defined and the closed-form solution is exactly what it claims to be. There are no load-bearing self-citations and no imported uniqueness theorems; the only self-referential element is the validation of the closed form against the same summation, which does not make the derivation circular because the analytical result is not being used to define the numerical result or vice versa.

Assumptions & free parameters 1 free parameters · 4 assumptions · 0 invented entities

The central derivation introduces no fitted constants; the only free parameter is the user-chosen locality scale sigma, and the falloff kernel itself is an ad hoc design choice. The load-bearing unproven assumptions are equivariance plus injectivity of the resulting plane map, with injectivity explicitly open.

free parameters (1)
  • sigma (falloff exponent) = user-controlled (examples use 1, 5, 10, 20)
    Controls locality of the deformation. The kernel c = 1 / 2^(sigma|x-n|+sigma|y-m|) is chosen so that the infinite lattice sum is a geometric series; the specific exponential-decay form is an ad hoc modeling choice, and sigma tunes its width.
assumptions (4)
  • domain assumption The displacement field is equivariant under the wallpaper group: u(g(x)) = g(u(x)) for every symmetry operation g.
    Stated in Section 4 as the starting observation. It ensures the deformed pattern retains the original symmetry, but does not by itself guarantee tileability.
  • domain assumption The map x -> x + u(x) is injective (a homeomorphism of the plane) for the deformations produced, so the warped tiling has no gaps or overlaps.
    Required for the headline 'no gaps or overlaps' claim. Not proven and explicitly deferred to future work in Section 6 ('ensure injectivity of the map under extreme deformations').
  • standard math The four Bravais lattice types (tetragonal, orthorhombic, hexagonal, monoclinic) suffice to express the guidance fields for all 17 wallpaper groups.
    Based on standard crystallography; the paper replaces the centered-rectangle lattice with parallelogram lattices 'of two directions' (Section 2), which is a reasonable but nonstandard re-framing.
  • ad hoc to paper The chosen falloff kernel produces visually meaningful and artistically controllable deformations.
    The kernel is selected for analytic summability, not derived from deformation principles; whether it yields desirable edits is validated only subjectively through examples.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Escher Tile Deformation via Closed-Form Solution." pith.science (2026). https://pith.science/paper/LCIOAXKO

@misc{pith2026250623388,
  author       = {Pith},
  title        = {Pith review of: Escher Tile Deformation via Closed-Form Solution},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/LCIOAXKO}},
  note         = {Machine review of arXiv:2506.23388}
}
read the original abstract

We present a real-time deformation method for Escher tiles -- interlocking organic forms that seamlessly tessellate the plane following symmetry rules. We formulate the problem as determining a periodic displacement field. The goal is to deform Escher tiles without introducing gaps or overlaps. The resulting displacement field is obtained in closed form by an analytical solution. Our method processes tiles of 17 wallpaper groups across various representations such as images and meshes. Rather than treating tiles as mere boundaries, we consider them as textured shapes, ensuring that both the boundary and interior deform simultaneously. To enable fine-grained artistic input, our interactive tool features a user-controllable adaptive fall-off parameter, allowing precise adjustment of locality and supporting deformations with meaningful semantic control. We demonstrate the effectiveness of our method through various examples, including photo editing and shape sculpting, showing its use in applications such as fabrication and animation.

Figures

Figures reproduced from arXiv: 2506.23388 by the authors.

Figure 1
Figure 1. Deforming both the boundary and interior of tiles via closed-form solution: Examples of the proposed real-time interactive deformation method. Tileability is preserved before and after deformation. Second row: All M.C. Escher works © 2025 The M.C. Escher Company - the Netherlands. All rights reserved. Used by permission. www.mcescher.com We present a real-time deformation method for Escher tiles—interlocking organic… view at source ↗
Figure 3
Figure 3. Comparison with naive point handle-based deformation: Edit￾ing the iconic gecko tile with the proposed tool. First column: before in￾teractive editing. Second column: ARAP after editing. Third column: the proposed method after editing. All M.C. Escher works © 2025 The M.C. Escher Company - the Netherlands. All rights reserved. Used by permission. www.mcescher.com different vertices of a regular grid. Since the guida… view at source ↗
Figure 2
Figure 2. Deformation with symmetry constraints: A monster with muscu￾lar back. A ballet dancer bending knee. A heron straightening neck. First row: All M.C. Escher works © 2025 The M.C. Escher Company - the Netherlands. All rights reserved. Used by permission. www.mcescher.com rotational or reflective transformation. At first sight, the problem is not even well-defined. In this work, we present an analytical method to deform… view at source ↗
Figures from the paper (11 more)
Figure 4
Figure 4. Figure 4: “Infinite Clock": The boundary of a tile is like an infinite clock. To remain a valid tile (that is the ability to cover the entire plane with no gap or overlap), when one boundary is extruding, there must be another boundary intruding. As can be observed from the disp…
Figure 5
Figure 5. Figure 5: Comparison: We validate the accuracy of our closed-form solution by comparing it with numerical summation. The coolwarm heatmap visual￾izes the analytical solution within a single cycle, accounting for infinitely many handle copies, while the viridis heatmap represents…
Figure 6
Figure 6. Figure 6: Deformable tiles for animation: Artists can create diverse facial expressions while keeping the lamp a valid tile [PITH_FULL_IMAGE:figures/full_fig_p006_6.png]
Figure 7
Figure 7. Figure 7: Deforming tiles of orthorhombic lattice: Deformation of dog tile of symmetry group ××. The solid circles specify the deformation near the user-input handle. The dashed circles specify the deformation caused by the infinite clock so that the resulted shape remains a val…
Figure 8
Figure 8. Figure 8: Deforming tiles of hexagonal lattice: Deformation of gingerbread tile of symmetry group 333. This example shows that the user wants to edit the leg of the gingerbread so that it looks like he is running. Because the puzzle is like an “infinite clock", when the leg defo…
Figure 9
Figure 9. Figure 9: Deforming tiles of monoclinic lattice: Deformation of seahorse tile of symmetry group 2222. This example shows that the user wants to edit the tail of the seahorse. Because the puzzle is like an “infinite clock", when the tail deforms, the wing deforms with it for the …
Figure 10
Figure 10. Figure 10: User interface of the photo editing tool: A point handle can be placed on the tile preview window at the lower left corner of the screen. The entire pattern in the photo will deform in real time according to the drag in the preview window. In this example, the user wa…
Figure 11
Figure 11. Figure 11: Locality control: Users can alter how local or global they wish for the brush by adjusting 𝜎. The extruding of beak corresponds to intruding of the back. Larger 𝜎 corresponds to more localized deformation near the push. Smaller 𝜎 corresponds to more global deformation…
Figure 13
Figure 13. Figure 13: Deformable tiles for fabrication: The pattern is deformed from top left to top bottom. Both before and after patterns are applied in the design of the dress [PITH_FULL_IMAGE:figures/full_fig_p009_13.png]
Figure 14
Figure 14. Figure 14: Deforming a packable cow into a packable rhino: This example shows the possibility to sculpt a volume from one animal to another while maintaining the shape as a valid tile. overlaps. Our closed-form solution enables simultaneous deforma￾tion of tile boundaries and in…
Figure 15
Figure 15. Figure 15: Editing a tile represented by a 2D triangle mesh: Enlarging the eye of the dinosaur head tile using multiple point handles. Left: before defor￾mation. Right: after deformation. Input artwork adapted from © mrhighsky from Adobe Stock. ACKNOWLEDGMENTS We are grateful to…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

32 extracted references · 30 canonical work pages

  1. [4]

    William Thurston

    Citeseer, 109–116. William Thurston. 1978-1981. The Geometry and Topology of Three-Manifolds. Princeton University lecture notes. Francis Williams and Eugene Zhang

  2. [5]

    On the Complete Evaluation of Jacobi Theta Functions

    On the Complete Evaluation of Jacobi Theta Functions. arXiv:1503.01141 [math.NT] Rulin Chen, Ziqi Wang, Peng Song, and Bernd Bickel

  3. [8]

    ACM Trans

    Dense, Interlocking-Free and Scalable Spectral Packing of Generic 3D Objects. ACM Trans. Graph. 42, 4 (2023), 141–1. Fernando De Goes and Doug L James

  4. [9]

    ACM Transactions on Graphics (TOG) 36, 4 (2017), 1–11

    Regularized kelvinlets: sculpting brushes based on fundamental solutions of elasticity. ACM Transactions on Graphics (TOG) 36, 4 (2017), 1–11. Fernando De Goes and Doug L James

  5. [16]

    In Proceedings of Bridges 2015: Mathematics, Music, Art, Architecture, Culture

    Algorithms for morphing escher-like tessellations. In Proceedings of Bridges 2015: Mathematics, Music, Art, Architecture, Culture . 483–486. Shih-Syun Lin, Charles C Morace, Chao-Hung Lin, Li-Fong Hsu, and Tong-Yee Lee

  6. [17]

    IEEE transactions on visualization and computer graphics 24, 2 (2017), 1103–1113

    Generation of escher arts with dual perception. IEEE transactions on visualization and computer graphics 24, 2 (2017), 1103–1113. Yuichi Nagata and Shinji Imahori

  7. [19]

    ACM Transactions on Graphics (TOG) 40, 3 (2021), 1–34

    Self-similar fractal drawings inspired by MC Escher’s print Square Limit. ACM Transactions on Graphics (TOG) 40, 3 (2021), 1–34. Peichang Ouyang and Robert W Fathauer

  8. [21]

    ACM Transactions on Graphics 42, 2 (2022), 1–17

    Interlocking spiral drawings inspired by MC Escher’s print whirlpools. ACM Transactions on Graphics 42, 2 (2022), 1–17. Jonathan Palacios and Eugene Zhang

Show all 32 references
  1. [25]

    IEEE Transactions on Visualization and Computer Graphics 24, 1 (2017), 843–852

    Interactive design and visualization of branched covering spaces. IEEE Transactions on Visualization and Computer Graphics 24, 1 (2017), 843–852. Doris Schattschneider

  2. [27]

    In ACM SIGGRAPH 2024 Conference Papers

    Into the Portal: Directable Fractal Self-Similarity. In ACM SIGGRAPH 2024 Conference Papers . 1–8. David Smith, Joseph Samuel Myers, Craig S Kaplan, and Chaim Goodman-Strauss

  3. [28]

    arXiv preprint arXiv:2303.10798 (2023)

    An aperiodic monotile. arXiv preprint arXiv:2303.10798 (2023). Walter F. Smith

  4. [29]

    https://www.physicssongs.org/ bravais.htm Accessed: 2024-07-07

    The Bravais Lattices Song. https://www.physicssongs.org/ bravais.htm Accessed: 2024-07-07. Olga Sorkine and Marc Alexa

  5. [31]

    In SIGGRAPH ASIA 2016 Technical Briefs (Macau) (SA ’16)

    Rendering kaleidoscopic scenes using orbifold theory. In SIGGRAPH ASIA 2016 Technical Briefs (Macau) (SA ’16) . As- sociation for Computing Machinery, New York, NY, USA, Article 17, 4 pages. https://doi.org/10.1145/3005358.3005368 Michael T Wong, Douglas E Zongker, and David H Salesin

  6. [32]

    Interactive Design and Optics-Based Visualization of Arbitrary Non-Euclidean Kaleidoscopic Orbifolds.IEEE Transactions on Visualization and Computer Graphics 30, 1 (2023), 1292–1301. , Vol. 1, No. 1, Article . Publication date: September 2025

  7. [1980]

    Doris Schattschneider

    Will it tile? Try the Conway criterion! Mathematics Magazine 53, 4 (1980), 224–233. Doris Schattschneider

  8. [1991]

    Mathematics Magazine 64, 4 (1991), 242–246

    Napoleon, Escher, and tessellations. Mathematics Magazine 64, 4 (1991), 242–246. Lawrence Roy, Prashant Kumar, Sanaz Golbabaei, Yue Zhang, and Eugene Zhang

  9. [2000]

    In Proceedings of the 27th annual conference on Computer graphics and interactive techniques

    Escherization. In Proceedings of the 27th annual conference on Computer graphics and interactive techniques . 499–510. Craig S Kaplan and David H Salesin. 2004b. Islamic star patterns in absolute geometry. ACM Transactions on Graphics (TOG) 23, 2 (2004), 97–119. Kevin Lee

  10. [2002]

    IEEE/ASME transactions on mechatronics 7, 4 (2002), 462–472

    Self-replicating robots for lunar development. IEEE/ASME transactions on mechatronics 7, 4 (2002), 462–472. John H Conway, Heidi Burgiel, and Chaim Goodman-Strauss

  11. [2006]

    ACM Transactions on Graphics (TOG) 25, 4 (2006), 1460–

    Periodic global parameterization. ACM Transactions on Graphics (TOG) 25, 4 (2006), 1460–

  12. [2007]

    ACM Transactions on Graphics (TOG) 26, 3 (2007), 55–es

    Rotational symmetry field design on surfaces. ACM Transactions on Graphics (TOG) 26, 3 (2007), 55–es. Nhan Phan-Thien and Sangtae Kim

  13. [2010]

    InProceedings of Bridges 2010: Mathematics, Music, Art, Architecture, Culture

    Curve evolution schemes for parquet deformations. InProceedings of Bridges 2010: Mathematics, Music, Art, Architecture, Culture . 95–102. Craig S. Kaplan and D. Salesin. 2004a. Dihedral Escherization. In Graphics Interface. https://api.semanticscholar.org/CorpusID:18798031 Cra...

  14. [2011]

    ACM Trans

    Bounded biharmonic weights for real-time deformation. ACM Trans. Graph. 30, 4 (2011),

  15. [2014]

    IEEE Computer Graphics and applications 34, 1 (2014), 68–76

    Beautiful math, part 2: aesthetic patterns based on fractal tilings. IEEE Computer Graphics and applications 34, 1 (2014), 68–76. Peichang Ouyang, Krzysztof Gdawiec, Alain Nicolas, David Bailey, and Kwok Wai Chung

  16. [2015]

    ACM Trans

    Orbifold tutte embeddings. ACM Trans. Graph. 34, 6 (2015), 190–1. Noam Aigerman and Yaron Lipman

  17. [2016]

    ACM Trans

    Hyperbolic orbifold tutte embeddings. ACM Trans. Graph. 35, 6 (2016), 217–1. Naum Ilich Akhiezer

  18. [2017]

    ACM Trans

    Spherical orbifold tu e embeddings. ACM Trans. Graph. 36, 4 (2017),

  19. [2018]

    ACM Transactions on Graphics (TOG) 37, 4 (2018), 1–10

    Dynamic Kelvinlets: secondary motions based on fundamental solutions of elastodynamics. ACM Transactions on Graphics (TOG) 37, 4 (2018), 1–10. Tom Goertzen

  20. [2020]

    Journal of Mathematics and the Arts 14, 1-2 (2020), 69–71

    KUHKUBUS-3D-Escher-tessellations. Journal of Mathematics and the Arts 14, 1-2 (2020), 69–71. Alec Jacobson, Ilya Baran, Jovan Popovic, and Olga Sorkine

  21. [2021]

    ACM Transactions on Graphics (TOG) 41, 2 (2021), 1–16

    Escherization with Large Deformations Based on As-Rigid-As-Possible Shape Modeling. ACM Transactions on Graphics (TOG) 41, 2 (2021), 1–16. Peichang Ouyang, Kwok Wai Chung, Alain Nicolas, and Krzysztof Gdawiec

  22. [2022]

    ACM Transactions on Graphics (TOG) 41, 4 (2022), 1–15

    Computational design of high-level interlocking puzzles. ACM Transactions on Graphics (TOG) 41, 4 (2022), 1–15. Gregory S Chirikjian, Yu Zhou, and Jackrit Suthakorn

  23. [2023]

    arXiv preprint arXiv:2309.14564 (2023)

    Generative Escher Meshes. arXiv preprint arXiv:2309.14564 (2023). Noam Aigerman, Shahar Z Kovalsky, and Yaron Lipman

  24. [2024]

    arXiv preprint arXiv:2405.15080 (2024)

    Constructing interlocking assemblies with crystallographic sym- metries. arXiv preprint arXiv:2405.15080 (2024). Branko Grünbaum and Geoffrey Colin Shephard

Pith tools

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