Pith. sign in

REVIEW 4 major objections 6 minor 37 references

Mechanics Simulation with Implicit Neural Representations of Complex Geometries

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

Pith's one-line read Finite element analysis can run directly on neural signed-distance-field geometry, using the Shifted Boundary Method to skip the explicit mesh and keep preprocessing time independent of triangle count.

desk verdict Promising INR–SBM integration that lacks a full-loop convergence study and carries a load-bearing sign inconsistency in the distance-vector formula. read the letter →

arxiv 2507.03087 v1 pith:65PSXECT submitted 2025-07-03 cs.CE

classification cs.CE
keywords ImplicitNeuralRepresentationsSignedDistanceFieldsShiftedBoundaryMethodLinearElasticityMesh-freesimulationOctreemeshesFiniteelementanalysisComplexgeometry
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

Finite element analysis usually needs an explicit surface mesh, but this paper proposes to run linear elasticity simulations directly on a neural network that encodes the geometry as a signed distance field. The key move is to pair the neural field with the Shifted Boundary Method (SBM), which applies boundary conditions on a grid-aligned surrogate boundary instead of the true surface. The network supplies the surrogate boundary, the inside/outside classification, and the distance vectors SBM requires, so no triangle soup or boundary-fitted mesh is needed. The authors demonstrate the workflow on a 2D ring benchmark, a 3D icosphere, and complex shapes from point clouds and triangle soups, reporting that meshing and assembly time no longer grow with the number of surface triangles.

What carries the argument

The central object is the neural signed distance field $f_\theta$, trained so that its gradient has unit norm near the boundary. SBM then uses the distance vector $d$ in a first-order shifted boundary condition on a surrogate octree boundary, enforced weakly through a consistency plus penalty term. The octree is built by querying the network for inside/outside classification of integration points, and distance vectors at the surrogate boundary are obtained by numerical gradient of the network, with a cache so each query is evaluated once. The same network therefore replaces both the mesh traversal and the closest-point projection that triangle-soup SBM would otherwise perform.

What would settle it

Take a high-curvature geometry such as the tetrakis or bunny, compute exact closest-point distance vectors at the integration points on the surrogate boundary, and compare against the network's gradient-derived vectors; if the cosine-similarity error grows beyond a few percent at sharp edges, the shifted boundary condition will misplace the displacement data there. A complementary test: run a mesh-convergence study with the INR in the loop—if the $L^2$ error stops decreasing at the level of the INR's distance-vector error (around $10^{-4}$), then the neural representation, not the SBM discretization, is the accuracy bottleneck.

Watch

Extended reading notes

Core claim

The central claim is that a neural signed distance field, trained with hybrid sampling near the boundary and a clamped, eikonal-regularized loss, can act as the geometric oracle for SBM end to end. For any point on the surrogate boundary, the network gives the signed distance, the unit gradient, and hence the distance vector, which is inserted into the shifted boundary condition that moves the displacement data from the true boundary to the surrogate boundary. The distance vector is taken as $d = -f_\theta(x)\nabla_x f_\theta(x)$, relying on the eikonal condition $\|\nabla_x f_\theta(x)\|=1$ near the surface. On the 2D ring with an analytical solution, the method recovers second-order convergence; on an icosphere, the surface-integrated displacement error against the triangle-soup ground truth is about $2.99\times 10^{-4}$. Wall-time experiments show constant meshing and assembly time for the INR as the triangle count grows, whereas the mesh-based pipeline scales with the number of triangles.

Load-bearing premise

The method stands or falls on whether the learned neural field behaves like a true signed distance function very close to the true boundary, so that its gradient—not just its value—points along the shortest path to the surface and gives the correct distance vector for the shifted boundary condition.

Editorial extensions

If this is right

  • Meshing and assembly wall times become independent of the number of surface triangles, so very large triangle soups can be analyzed at constant preprocessing cost.
  • Geometries from point clouds, multi-view images, or generative models can be simulated without converting them to a boundary-fitted mesh.
  • Adaptive octree refinement near boundaries can be driven directly by neural queries, automatically resolving fine geometric features.
  • The framework currently applies to linear elasticity and linear PDEs; the authors propose extending it to large-deformation and multi-physics problems.

Reading between the lines

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

  • An unstated consequence of the wall-time plot is that the INR advantage depends on the network size and hardware: for very small triangle counts, the fixed cost of the neural network forward pass could exceed the cost of traversing a small triangle soup.
  • The paper's own error maps show the largest distance-vector errors at sharp curvature, which suggests that CAD-style geometry with edges and corners will need curvature-aware training or adaptive sampling before the method reaches engineering accuracy there.
  • The distance-vector definition in Eq. (9) and the sign used in Algorithm 4 are opposite; an implementation must settle the sign convention consistently, otherwise the shifted boundary condition would push the displacement data in the wrong direction.
  • A direct testable extension is to repeat the ring convergence study with the INR supplying the boundary: the convergence slope should flatten at the mesh size where the neural distance-vector error starts to dominate, giving a practical accuracy ceiling for the approach.
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. The manuscript proposes a computational pipeline that couples implicit neural representations (INRs) of geometry with the shifted boundary method (SBM) for linear elasticity. The authors train an MLP-based signed distance field from a triangle soup using a hybrid sampling strategy and a loss that combines a clamped distance term with eikonal and normal-similarity regularization. The trained INR is then used to classify octree cells, construct an incomplete octree, identify the surrogate boundary, and compute distance vectors for SBM's shifted Taylor boundary condition. The paper validates the SBM solver on a ring benchmark with a known analytical solution, compares an INR-based icosphere against its triangle-mesh counterpart in one configuration, and presents simulation results for the Stanford Bunny, the Eiffel Tower, and a gyroid.

Significance. If substantiated, the framework would remove the explicit surface triangulation from the simulation pipeline and make boundary-related geometric queries independent of the surface triangle count, a practically useful integration of INR geometry with embedded finite element analysis. The paper has several strengths: the SBM ring convergence test is a sound check of the solver; the proposed loss and sampling strategy are clearly specified; and the wall-time comparison in Figure 11 directly demonstrates the expected asymptotic advantage of neural queries over triangle-soup traversal. However, the central claim that INR geometry can replace the meshing step for accurate SBM analysis is not yet fully supported: the only INR-in-the-loop accuracy test is a single fixed-refinement icosphere case, the distance-vector sign convention is internally inconsistent, and Section 5.1 documents substantial distance-vector misalignment near sharp features. These are load-bearing gaps because SBM consistency relies on the distance vector being an accurate closest-point vector of size O(h).

major comments (4)
  1. [Section 2.3, Eq. (8)-(9); Algorithm 3; Algorithm 4] The sign convention for the signed distance field is internally inconsistent. Section 2.3, Eq. (8) defines f > 0 inside the domain and f < 0 outside, and Eq. (9) gives d = -f_theta grad f_theta. In contrast, Algorithm 3 classifies a Gauss point as interior when f_theta(gp) < 0, and Algorithm 4 computes d_gp = (grad f_theta / ||grad f_theta||) f_theta(Q). Under the Eq. (8) convention, Eq. (9) points from an interior point toward the interior, opposite to the closest-point projection to the true boundary; under the Algorithm 3 convention, Algorithm 4 points inward rather than outward. The discrepancy is not cosmetic because the shifted Taylor condition in Eq. (20) requires d = x - x-tilde to point from the surrogate boundary to the true boundary. Please correct the sign convention consistently across Eq. (8), Eq. (9), Algorithm 3, and Algorithm 4, and state the chosen convention explicitly.
  2. [Sections 5.2.1-5.2.2] The convergence study in Section 5.2.1 uses an analytic ring geometry and therefore validates the SBM formulation, not the INR-in-the-loop pipeline. The only direct INR comparison, Section 5.2.2, is a single icosphere at base level 4 and boundary level 8, reporting one surface L2 error of 2.99e-4 with no mesh refinement sequence, no repeated training runs, and no error bars. To support the claim that the INR-SBM combination achieves the expected second-order convergence, please report L2 displacement error versus mesh size for the INR-based icosphere (or another INR geometry) over several refinement levels, including at least three independently trained INRs to quantify training variability.
  3. [Section 5.1, Table 5] Table 5 reports a mean cosine similarity of the distance vector of 0.980 with standard deviation 0.130 for the turbine, and Figure 7 shows that misalignment concentrates at sharp edges. Because the consistency and adjoint-consistency terms in Eq. (24) rely on d = x - x-tilde with ||d|| = O(h), an O(1) angular error in a boundary-layer region can introduce an error in the Taylor remainder of Eq. (20) that does not vanish under mesh refinement. Please quantify the effect of the measured distance-vector errors on the SBM consistency, for example by comparing ring solutions with exact and INR-perturbed distance vectors, or by reporting the convergence of the INR distance-vector error with respect to h at the Gauss points used by the solver.
  4. [Abstract and Section 4.3] The abstract states that the framework 'effectively eliminates the meshing step,' but the pipeline still requires generating and adaptively refining an octree mesh; what is eliminated is the explicit surface triangulation as the geometric input. This distinction is important for assessing the claimed computational advantages, and the wording should be adjusted so that the claim is not overstated.
minor comments (6)
  1. [Section 4.1] The heading contains a typo: 'Dirihclet' should be 'Dirichlet.'
  2. [Section 4.2] The text contains a typo: 'exapnsion' should be 'expansion.'
  3. [Section 3, Eq. (10)] The loss function in Eq. (10) uses the ground-truth signed distance s and normal n-hat, but s is not defined immediately before Eq. (10); please state that s denotes the target signed distance from the triangle soup and n-hat denotes the true surface normal.
  4. [Section 5.1, Figures 6 and 7] The color bars in Figures 6 and 7 are labeled 'Order of Error' with values from about 4.5 to 8, while the text says the error is on the order of 10^-4; the figures appear to plot -log10(error) or a similar transformed quantity. Please relabel the color bars and clarify the plotted quantity.
  5. [Section 5.2.2, Figure 11] The wall-time comparison in Figure 11 does not specify whether the octree refinement levels are kept fixed as the number of surface triangles is increased; without this detail, the comparison may conflate surface-mesh refinement with volume-mesh refinement. Please state the fixed setup or explain how the triangle count is varied.
  6. [Algorithm 4] The statement that the gradient is computed numerically 'by using two stencils on each axis using the central difference method' is vague; please specify the stencil size and how it relates to the mesh size h.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the INR-SBM pipeline is an accuracy-checked integration of existing SBM with a learned geometric oracle, not a derivation that reduces to its inputs.

full rationale

The paper's chain is: train an INR to approximate the signed distance field of a triangle soup; query the INR for the surrogate boundary (zero level set) and distance vector (Eq. 9, the standard SDF identity d = -f grad f); feed these into the existing Shifted Boundary Method; and compare the resulting displacements against analytic or mesh-based references. None of these steps fits the target physics output or defines it in terms of the method. The INR is trained on geometric ground truth (SDF values and normals from libigl), and its fidelity is then measured at Gauss points against the same type of ground truth (Section 5.1 and Table 5); that is an approximation-error estimate, not a circular derivation. The 2D ring benchmark (Section 5.2.1) checks SBM against an analytic solution, and the icosphere study (Section 5.2.2) compares INR-driven SBM with mesh-driven SBM on identical octree refinement, reporting a small surface error. Citations to the authors' prior work ([12,17-21]) supply the octree refinement and surrogate-boundary-selection machinery, but these are independently published algorithmic components rather than the paper's claimed result; no uniqueness theorem or fitted physical parameter is being re-imported as a prediction. The main weaknesses are non-circular correctness concerns: Eq. 9 and Algorithm 4 differ in sign convention, no INR-in-the-loop mesh-convergence study is run, and Section 5.1 shows distance-vector error concentrating at sharp curvature; these affect robustness and evidentiary strength, not circularity.

Assumptions & free parameters 7 free parameters · 5 assumptions · 0 invented entities

The contribution is an integration of two established methods rather than a derivation; the listed hyperparameters are chosen by hand or by Bayesian optimization and are not fitted to mechanical observables. No new physical entities, forces, or conserved quantities are introduced.

free parameters (7)
  • lambda_g (eikonal loss weight) = not reported
    Introduced in Eq. 10 as a multiplier for the eikonal constraint; the ablation in Table 3 shows removing the eikonal term degrades NMSE, so the value matters but is not given.
  • tau (normal similarity loss weight) = not reported
    Same Eq. 10; weight for aligning the INR gradient with true normals; no value is reported.
  • delta (narrow-band width) = 0.001 in experiments
    Defines the clamping band in the loss and the sampling region; chosen by hand and directly controls which points determine boundary behavior.
  • omega (geometric regularization threshold) = not reported
    Appears in Eq. 10 to limit eikonal/normal terms to |s| < omega; controls how close to the boundary regularization is applied, but is not quantified.
  • hybrid sampling counts n(P_U), n(P_S), n(P_NB) = 90K, 28K, 32K (Bayesian optimized)
    Table 4 shows the hybrid counts were tuned by Bayesian optimization to minimize NMSE; these counts are part of the method rather than derived from theory.
  • SBM Nitsche penalty gamma = not reported
    Required in Eqs. 23-24 for weak Dirichlet enforcement; no value or stabilization rule is given, making reproduction harder.
  • octree refinement levels (base/boundary) = e.g., 5/9 bunny, 6/10 tower, 6/8 gyroid
    Chosen per geometry without an automatic rule, despite the paper claiming automated adaptive refinement.
assumptions (5)
  • domain assumption SBM shifted boundary condition via Taylor expansion plus Nitsche's method converges to the true boundary condition as the distance vector tends to zero.
    Adopted from Main and Scovazzi [7,8] and Atallah et al. [17,33]; invoked in Section 4.1-4.2 without re-derivation.
  • domain assumption The neural network field approximates a signed distance function satisfying the Eikonal equation in the narrow band.
    Eq. 9 assumes ||grad f_theta|| = 1 and defines the distance vector from the gradient; Algorithm 4 queries this gradient as if it were the true closest-point direction. Section 5.1 shows this is only approximate, with larger errors at high curvature.
  • domain assumption Classification of grid points as inside/outside from the INR sign determines a correct surrogate domain.
    Algorithms 2 and 3 use f_theta(gp) < 0 as interior; any sign error near the surface changes the surrogate boundary and all downstream distance vectors.
  • standard math Linear elasticity constitutive law and small-strain assumption.
    Section 4 uses isotropic linear elasticity (Eqs. 12-14); the method is only demonstrated for this regime, not for large deformation or nonlinear materials.
  • domain assumption Octree 2:1 balancing preserves mesh quality and SBM accuracy.
    Adopted from Saurabh et al. [20]; used in the adaptive octree generation and refinement criteria without analysis in this paper.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Mechanics Simulation with Implicit Neural Representations of Complex Geometries." pith.science (2026). https://pith.science/paper/65PSXECT

@misc{pith2026250703087,
  author       = {Pith},
  title        = {Pith review of: Mechanics Simulation with Implicit Neural Representations of Complex Geometries},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/65PSXECT}},
  note         = {Machine review of arXiv:2507.03087}
}
read the original abstract

Implicit Neural Representations (INRs), characterized by neural network-encoded signed distance fields, provide a powerful means to represent complex geometries continuously and efficiently. While successful in computer vision and generative modeling, integrating INRs into computational analysis workflows, such as finite element simulations, remains underdeveloped. In this work, we propose a computational framework that seamlessly combines INRs with the Shifted Boundary Method (SBM) for high-fidelity linear elasticity simulations without explicit geometry transformations. By directly querying the neural implicit geometry, we obtain the surrogate boundaries and distance vectors essential for SBM, effectively eliminating the meshing step. We demonstrate the efficacy and robustness of our approach through elasticity simulations on complex geometries (Stanford Bunny, Eiffel Tower, gyroids) sourced from triangle soups and point clouds. Our method showcases significant computational advantages and accuracy, underscoring its potential in biomedical, geophysical, and advanced manufacturing applications.

Figures

Figures reproduced from arXiv: 2507.03087 by the authors.

Figure 1
Figure 1. Left, middle, and right represent the displacement magnitude in [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. The domain Ω is a square grid, with a circle at the center, featuring the true boundary Γ and the surrogate boundary Γ˜ . We present a method to generate analysis suitable INR to demon￾strate the training complexity involved for INR along with that to get INR with favorable ground truth described in Section 3. We show the formulation for linear elasticity along with the shifted boundary conditions in Section 4. In S… view at source ↗
Figure 3
Figure 3. Ico-Sphere uniformly sampled points, 25K points in narrowband with width 𝛿 = 0.001, and 25K points in the surface. In this study, we perform following comparisons: (1) Comparison of the Implicit Net with different loss functions along the modified Implicit Loss as given in Equation 10. (2) Ablation study of loss function in Equation 10. (3) Ablation study of the sampling strategy [PITH_FULL_IMAGE:figures/full_fig_p… view at source ↗
Figures from the paper (13 more)
Figure 4
Figure 4. Figure 4: Neural inference required for computing the distance vector occurs [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: shows the surrogate boundary and the gauss points in the surrogate boundary with the distance vector pointing towards the true boundary, we are basically measuring the correctness of this vector [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]
Figure 6
Figure 6. Figure 6: Plot of 𝑙𝑜𝑔10 ( |𝑓𝜃 (𝑥𝑔𝑝 ) − 𝑠 (𝑥𝑔𝑝 ) | ) for refinement ℎ = Δ/2 8 . The error mostly is in order of 10−4 for all the geometries. The plot shows the spatial variation of error in the magnitude of the distance vectors. (a) Sphere (b) Cylinder (c) Cone (d) Tetrakis (e) B…
Figure 7
Figure 7. Figure 7: Plot of 𝑙𝑜𝑔10 (1 − ⟨𝑑 true 𝑔𝑝 · 𝑑 𝑓𝜃 𝑔𝑝 ⟩ ) for refinement ℎ = Δ/2 8 . The error shows the magnitude of the misalignment of the distance vector. As is more evident in tetrakis, the error is very high in the edges where there is a sharp change in curvature. Overall the …
Figure 8
Figure 8. Figure 8: Top: Log-log plot showing 𝐿2 error convergence with mesh refinement. Right: Normalized 𝐿2 error as a bar chart across levels for a ring with Dirichlet boundary condition applied on both inner (using SBM) and outer ring (strongly) [PITH_FULL_IMAGE:figures/full_fig_p010…
Figure 9
Figure 9. Figure 9: Displacement magnitude of the full ring (background). Top-right [PITH_FULL_IMAGE:figures/full_fig_p010_9.png]
Figure 10
Figure 10. Figure 10: Solution field and corresponding error in the displacement field between ico-sphere and corresponding INR. The solution obtained from the octree-grids [PITH_FULL_IMAGE:figures/full_fig_p011_10.png]
Figure 11
Figure 11. Figure 11: Log-log plot of meshing and assembly wall times with increasing [PITH_FULL_IMAGE:figures/full_fig_p011_11.png]
Figure 12
Figure 12. Figure 12: (a) INR-based Stanford Bunny with a base refinement of level 5. [PITH_FULL_IMAGE:figures/full_fig_p011_12.png]
Figure 13
Figure 13. Figure 13: (a) Displacement magnitude ((𝑢 2 𝑥 + 𝑢 2 𝑦 + 𝑢 2 𝑧 ) 1/2 ) along a slice of the Bunny (shown in [PITH_FULL_IMAGE:figures/full_fig_p012_13.png]
Figure 15
Figure 15. Figure 15: (a) and (b) show the displacement 𝑈𝑥 and 𝑈𝑦, which varies sinu￾soidally as per the prescribed boundary condition. The displacement in the 𝑥-direction is higher. The tower is fixed in the 𝑧-direction. condition enforced using SBM everywhere in the boundary: 𝑢𝑥 = 0.1 si…
Figure 16
Figure 16. Figure 16: Displacement fields (a) 𝑈𝑥 and (b) 𝑈𝑦 under a radially increasing displacement boundary condition. Displacements are minimal near the center and increase outward. (c) Radial displacement magnitude for the Gyroid, confirming consistency with the applied boundary condit…
Figure 17
Figure 17. Figure 17: (a) Octree representation of complicated Gyroid shape obtained [PITH_FULL_IMAGE:figures/full_fig_p013_17.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

37 extracted references · 31 canonical work pages

  1. [1]

    McHenry, P

    K. McHenry, P. Bajcsy, An overview of 3D data content, file formats and viewers, Technical Report 1205, National Center for Supercomputing Applications (NCSA),

  2. [2]

    Chiba, I

    N. Chiba, I. Nishigaki, Y. Yamashita, C. Takizawa, K. Fujishiro, A flexible auto- matic hexahedral mesh generation by boundary-fit method, Computer Meth- ods in Applied Mechanics and Engineering 161 (1998) 145–154. URL: https: //www.sciencedirect.com/science/article/pii/S0045782597003046. doi:https:// doi.org/10.1016/S0045-7825(97)00304-6

  3. [3]

    J. J. Park, P. Florence, J. Straub, R. Newcombe, S. Lovegrove, Deepsdf: Learning continuous signed distance functions for shape representation, in: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2019, pp. 165–174

  4. [4]

    Gropp, L

    A. Gropp, L. Yariv, N. Haim, M. Atzmon, Y. Lipman, Implicit geometric regular- ization for learning shapes, arXiv preprint arXiv:2002.10099 (2020)

  5. [5]

    Mescheder, M

    L. Mescheder, M. Oechsle, M. Niemeyer, S. Nowozin, A. Geiger, Occupancy networks: Learning 3d reconstruction in function space, in: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2019, pp. 4460–4470

  6. [6]

    Z. Chen, H. Zhang, Learning implicit fields for generative shape modeling, in: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2019, pp. 5939–5948

  7. [7]

    A. Main, G. Scovazzi, The shifted boundary method for embedded domain computations. part i: Poisson and stokes problems, Journal of Computational Physics 372 (2018) 972–995

  8. [8]

    A. Main, G. Scovazzi, The shifted boundary method for embedded domain computations. part ii: Linear advection–diffusion and incompressible navier– stokes equations, Journal of Computational Physics 372 (2018) 996–1026

Show all 37 references
  1. [9]

    M.-C. Hsu, C. Wang, F. Xu, A. J. Herrema, A. Krishnamurthy, Direct immersogeo- metric fluid flow analysis using b-rep cad models, Computer Aided Geometric Design 43 (2016) 143–158

  2. [10]

    Burman, S

    E. Burman, S. Claus, P. Hansbo, M. G. Larson, A. Massing, Cutfem: discretizing geometry and partial differential equations, International Journal for Numerical Methods in Engineering 104 (2015) 472–501

  3. [11]

    Burman, P

    E. Burman, P. Hansbo, Fictitious domain finite element methods using cut ele- ments: Ii. a stabilized nitsche method, Applied Numerical Mathematics 62 (2012) 328–341

  4. [12]

    Saurabh, B

    K. Saurabh, B. Gao, M. Fernando, S. Xu, M. A. Khanwale, B. Khara, M.-C. Hsu, A. Krishnamurthy, H. Sundar, B. Ganapathysubramanian, Industrial scale large eddy simulations with adaptive octree meshes using immersogeometric analysis, Computers & Mathematics with Applications 97 ...

  5. [13]

    Colomés, A

    O. Colomés, A. Main, L. Nouveau, G. Scovazzi, A weighted shifted boundary method for free surface flow problems, Journal of Computational Physics 424 (2021) 109837

  6. [14]

    G. Chou, Y. Bahat, F. Heide, Diffusion-sdf: Conditional generative modeling of signed distance functions, in: Proceedings of the IEEE/CVF international conference on computer vision, 2023, pp. 2262–2272

  7. [15]

    Huang, H.-C

    T. Huang, H.-C. Lim, Simulation of lid-driven cavity flow with internal circular obstacles, Applied Sciences 10 (2020) 4583

  8. [16]

    E. N. Karatzas, G. Stabile, L. Nouveau, G. Scovazzi, G. Rozza, A reduced-order shifted boundary method for parametrized incompressible navier–stokes equa- tions, Computer Methods in Applied Mechanics and Engineering 370 (2020) 113273

  9. [17]

    N. M. Atallah, C. Canuto, G. Scovazzi, The shifted boundary method for solid mechanics, International Journal for Numerical Methods in Engineering 122 (2021) 5935–5970

  10. [18]

    C.-H. Yang, K. Saurabh, G. Scovazzi, C. Canuto, A. Krishnamurthy, B. Ganapa- thysubramanian, Optimal surrogate boundary selection and scalability studies for the shifted boundary method on octree meshes, Computer Methods in Applied Mechanics and Engineering 419 (2024) 116686

  11. [19]

    C.-H. Yang, G. Scovazzi, A. Krishnamurthy, B. Ganapathysubramanian, Simulating incompressible flows over complex geometries using the shifted boundary method with incomplete adaptive octree meshes, arXiv preprint arXiv:2411.00272 (2024)

  12. [20]

    Saurabh, M

    K. Saurabh, M. Ishii, M. Fernando, B. Gao, K. Tan, M.-C. Hsu, A. Krishnamurthy, H. Sundar, B. Ganapathysubramanian, Scalable adaptive pde solvers in arbitrary domains, in: Proceedings of the International Conference for high performance computing, networking, storage and analy...

  13. [21]

    Jignasu, E

    A. Jignasu, E. Herron, Z. Jiang, S. Sarkar, C. Hegde, B. Ganapathysubramanian, A. Balu, A. Krishnamurthy, Stitch: Surface reconstruction using implicit neu- ral representations with topology constraints and persistent homology, arXiv preprint arXiv:2412.18696 (2024)

  14. [22]

    S. Liu, S. Saito, W. Chen, H. Li, Learning to infer implicit surfaces without 3d supervision, Advances in Neural Information Processing Systems 32 (2019)

  15. [23]

    Ben-Shabat, C

    Y. Ben-Shabat, C. H. Koneputugodage, S. Gould, Digs: Divergence guided shape implicit neural representation for unoriented point clouds, in: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2022, pp. 19323–19332

  16. [24]

    Atzmon, Y

    M. Atzmon, Y. Lipman, Sal: Sign agnostic learning of shapes from raw data, in: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2020, pp. 2565–2574

  17. [25]

    P. Wang, L. Liu, Y. Liu, C. Theobalt, T. Komura, W. Wang, Neus: Learning neural implicit surfaces by volume rendering for multi-view reconstruction, arXiv preprint arXiv:2106.10689 (2021)

  18. [26]

    Niemeyer, L

    M. Niemeyer, L. Mescheder, M. Oechsle, A. Geiger, Differentiable volumetric rendering: Learning implicit 3d representations without 3d supervision, in: Pro- ceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2020, pp. 3504–3515

  19. [27]

    J. Ho, A. Jain, P. Abbeel, Denoising diffusion probabilistic models, Advances in neural information processing systems 33 (2020) 6840–6851

  20. [28]

    Rombach, A

    R. Rombach, A. Blattmann, D. Lorenz, P. Esser, B. Ommer, High-resolution image synthesis with latent diffusion models, in: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2022, pp. 10684–10695

  21. [29]

    Erkoç, F

    Z. Erkoç, F. Ma, Q. Shan, M. Nießner, A. Dai, Hyperdiffusion: Generating implicit neural fields with weight-space diffusion, in: Proceedings of the IEEE/CVF international conference on computer vision, 2023, pp. 14300–14310

  22. [30]

    Sitzmann, J

    V. Sitzmann, J. Martel, A. Bergman, D. Lindell, G. Wetzstein, Implicit neural rep- resentations with periodic activation functions, Advances in neural information processing systems 33 (2020) 7462–7473

  23. [31]

    Jacobson, D

    A. Jacobson, D. Panozzo, et al., libigl: A simple C++ geometry processing library,

  24. [32]

    Schillinger, M

    D. Schillinger, M. Ruess, N. Zander, Y. Bazilevs, A. Düster, E. Rank, Small and large deformation analysis with the p-and b-spline versions of the finite cell method, Computational Mechanics 50 (2012) 445–478

  25. [33]

    N. M. Atallah, C. Canuto, G. Scovazzi, The second-generation shifted boundary method and its numerical analysis, Computer Methods in Applied Mechanics and Engineering 372 (2020) 113341

  26. [34]

    G. Turk, M. Levoy, Zippered polygon meshes from range images, in: Proceedings of the 21st annual conference on Computer graphics and interactive techniques, 1994, pp. 311–318

  27. [35]

    A. H. Schoen, Infinite periodic minimal surfaces without self-intersections, vol- ume 5541, National Aeronautics and Space Administration, 1970

  28. [2008]

    Technical Report, 22 pages

  29. [2018]

    Mechanics Simulation with Implicit Neural Representations of Complex Geometries•15

    Https://libigl.github.io/. Mechanics Simulation with Implicit Neural Representations of Complex Geometries•15

Pith tools

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