Open-Source & Code

Public implementations released with peer-reviewed work so others can reproduce, extend, and compare methods fairly.

Releasing code alongside papers reduces the gap between published algorithms and what practitioners can actually run. Where possible, I publish repositories for FastSpectrum (fast Laplace–Beltrami eigen approximation), LocalFields (locally supported tangential fields), and collaborate on releases such as DeltaConv and GravoMG (geometric multigrid on surfaces).

These projects are maintained in the spirit of academic reproducibility: cite the corresponding paper if you build on the implementation, open issues for bugs, and consider contributing improvements back when your fork stabilises.

FastSpectrum — fast Laplace–Beltrami approximation

Reference implementation for the fast eigenproblem approximation framework described in the 2018 CGF / SGP paper. Useful when you need approximate spectral data on meshes without paying for a full dense or iterative solve at interactive rates.

Related publications

LocalFields — tangential vector fields on surfaces

Code for constructing and editing locally supported direction fields and related representations on triangle meshes, as in the Eurographics 2020 CGF paper.

Related publications

DeltaConv — geometric deep learning on point clouds

Training code and models for anisotropic point-cloud convolutions; accompanies the ACM TOG 2022 publication and the earlier arXiv release.

Related publications

GravoMG — geometric multigrid for curved surfaces

Implementation associated with the SIGGRAPH 2023 paper on fast geometric multigrid for surface PDEs; suitable for researchers comparing solver performance on mesh hierarchies.

Related publications

Additional publications with formal artefacts

The hierarchical subspace iteration method is documented in ACM TOG with DOI links; while the reference code path may differ from the repositories above, the paper remains the authoritative specification for that algorithm.

Related publications

Publications

All papers tied to released code

Open-source or strongly reproducibility-oriented entries; newest first.

2023

A Fast Geometric Multigrid Method for Curved Surfaces

Wiersma, R., Nasikun, A., Eisemann, E., & Hildebrandt, K.

ACM SIGGRAPH 2023 Conference Proceedings, pp. 1–11 · ACM

2022

DeltaConv: Anisotropic Operators for Geometric Deep Learning on Point Clouds

Wiersma, R., Nasikun, A., Eisemann, E., & Hildebrandt, K.

ACM Transactions on Graphics (TOG), 41(4), pp. 1–10 · ACM

2022

The Hierarchical Subspace Iteration Method for Laplace–Beltrami Eigenproblems

Nasikun, A., & Hildebrandt, K.

ACM Transactions on Graphics (TOG), 41(2), pp. 1–14 · ACM

2022

DeltaConv: Anisotropic Geometric Deep Learning with Exterior Calculus

Wiersma, R., NASIKUN, A., Eisemann, E., & Hildebrandt, K.

Additional Scholar listing (related to DeltaConv / TOG 2022), 2022

2021

DeltaConv: Anisotropic Point Cloud Learning with Exterior Calculus

Wiersma, R., Nasikun, A., Eisemann, E., & Hildebrandt, K.

arXiv preprint arXiv:2111.08799, 2021

2020

Locally Supported Tangential Vector, n-Vector, and Tensor Fields

Nasikun, A., Brandt, C., & Hildebrandt, K.

Computer Graphics Forum, 39(2), pp. 203–217 · Eurographics 2020

2018

Fast Approximation of Laplace–Beltrami Eigenproblems

Nasikun, A., Brandt, C., & Hildebrandt, K.

Computer Graphics Forum, 37(5), pp. 121–134 · SGP 2018 · Eurographics

← Full publication list on homepage