Lightnews — Scholar-powered news

David Coeurjolly @dcoeurjo.bsky.social · 7d

Glad to share with you our last paper to be presented @ SIGGRAPH Asia 2025 !
Kudos @baptiste-genest.bsky.social

Baptiste GENEST @baptiste-genest.bsky.social · 7d

Computing the exact bijection of the optimal transport (OT) problem between very large point sets is completely untractable…

In our SIGGRAPH Asia 2025 paper: “BSP-OT: Sparse transport plans between discrete measures in log-linear time” we get one with typically 1% of error in a few seconds on CPU!

1 7

David Coeurjolly @dcoeurjo.bsky.social · Aug 23

New in 2.1 (Bastien Doignies, @cnrs.fr, LIRIS):

* 4 new samplers (from siggraph 2020, 2021, 2024 and 2025 papers)
* New metrics (semi-discrete Optimal Transport and Gaussian kernels)
* bugfixes..

Code: github.com/utk-team/utk
Release note: github.com/utk-team/utk...

GitHub - utk-team/utk: Uni{corn|form} toolkit

Uni{corn|form} toolkit . Contribute to utk-team/utk development by creating an account on GitHub.

github.com

1

David Coeurjolly @dcoeurjo.bsky.social · Aug 23

Glad to share with you the release 2.1 of the UniformToolKit utk-team.github.io/utk/

UTK is a C++ and Python library to generate and analyze sampling patterns for numerical Monte-Carlo integration (eg. for path tracing algorithms).

1 5 28

David Coeurjolly @dcoeurjo.bsky.social · Aug 21

Timelapse (beginning only)

3

David Coeurjolly @dcoeurjo.bsky.social · Aug 21

Axis plotter again... Voxelized Nerfertiti + random block shifts.

More here: perso.liris.cnrs.fr/david.coeurj...
#axisplotter

1 1 7

David Coeurjolly @dcoeurjo.bsky.social · Jul 10

We have also completely refactored the cmake scripts and the dependencies management: no more GMP, Boost::Integer instead, no more Qt, deps fetched using cmake/CPM...

David Coeurjolly @dcoeurjo.bsky.social · Jul 10

and a brand new paraview (paraview.org) plugin to use DGtal features in paraview pipelines!

1 1

David Coeurjolly @dcoeurjo.bsky.social · Jul 10

This version also contains many new features (see github.com/DGtal-team/D...). Among them, new corrected normal current differential estimators and volumetric tools (Voronoi maps and distance transformation) in the Shortcuts

1 1

David Coeurjolly @dcoeurjo.bsky.social · Jul 10

Huge improvements on the python binding with many Shortcuts available 🐍(www.dgtal.org/doc/stable/m...)

1

David Coeurjolly @dcoeurjo.bsky.social · Jul 10

This new version includes a brand new viewer that is based on polyscope (polyscope.run).

1

David Coeurjolly @dcoeurjo.bsky.social · Jul 10

We are happy to announce the major release 2.0 of DGtal (dgtal.org) 🎉. This big update makes the library more modern, in both how it is built and the tools it uses. This release also includes new features in both the C++ library and its python binding. #geometry-processing #c++ #python (1/...)

1 2 8

David Coeurjolly @dcoeurjo.bsky.social · Jun 11

You can still use tricks. Eg fixing a max number of samples and optimising OT + progressivity for previous points, or use tile-based approaches projet.liris.cnrs.fr/polyhex/), or change the model/ energy (clamping in GBN or progressive Poisson disk), but you won’t get the same quality.

projet.liris.cnrs.fr

David Coeurjolly @dcoeurjo.bsky.social · Jun 11

Oh I see… I missed the pointset vs sequence thing.
So yes having BN sequences could be hard depending on your mathematical BN model you consider. Top quality ones (GBN or OT based) rely on a global energy which may by incompatible with progressively generating samples.

1

David Coeurjolly @dcoeurjo.bsky.social · Jun 9

www.geometry.caltech.edu/BlueNoise/ , abdallagafar.com/publications... or perso.liris.cnrs.fr/david.coeurj... ?
(more refs available in the related work sections)

Gaussian Blue Noise

abdallagafar.com

2 2

Reposted by David Coeurjolly

Emilie Yu @emxtyu.bsky.social · Jun 1

Ever wondered how badly we're all addicted to buying new GPUs in graphics labs?

Come see our talk at #SIGGRAPH2025 to discuss how we can collectively move "Towards a sustainable use of GPUs in Graphics Research"

with @elie-michel.bsky.social @axelparis.bsky.social Octave Crespel and Felix Hähnlein

MOTIVATION
Graphical Processing Units (GPUs) are at the core of Computer Graphics research. These chips are critical for rendering images, processing geometric data, and training machine learning models. Yet, the production and disposal of GPUs emits CO2 and results in toxic e-waste [1].

METHOD
We surveyed 888 papers presented at SIGGRAPH (premier conference for computer graphics research), from 2018 to 2024, and systematically gathered GPU models cited in the text.

We then contextualize the hardware reported in papers with publicly available data of consumers’ hardware [2, 3].

REFERENCES
[1] CRAWFORD, KATE. The Atlas of AI: Power, Politics, and the Planetary Costs of Artificial Intelligence. Yale University Press, 2021.
[2] STEAM. Steam Hardware Survey. https://store.steampowered.com/hwsurvey
[3] BLENDER. Blender Open Data. https://opendata.blender.org

2 20 52

Reposted by David Coeurjolly

Baptiste GENEST @baptiste-genest.bsky.social · May 10

Implicit surfaces are great, but if you are not Inigo Quilez, it's really hard to control how they look...

That's why I'm really proud to annonce our Eurographics 2025 paper: "Implicit UVs: Real-time semi-global parameterization of implicit surfaces".

4 15 54

David Coeurjolly @dcoeurjo.bsky.social · May 5

I couldn’t agree more, definitely a future award-winning movie!

2

David Coeurjolly @dcoeurjo.bsky.social · May 5

A second paper has been accepted to SIGGRAPH 2025! Great collaboration with Khoa Do, @nbonneel.bsky.social , and Pooran Memari

bsky.app/profile/nbon...

Nicolas Bonneel @nbonneel.bsky.social · May 5

A thread on our new, also awesome, Siggraph paper "Linear-Time Transport with Rectified Flows", by K. Do, @dcoeurjo.bsky.social, P. Memari and me.
Paper here: perso.liris.cnrs.fr/nicolas.bonn...
Results in video: youtu.be/EcfKnSe6mhk

Linear-Time Transport with Rectified Flows

YouTube video by David Coeurjolly

youtu.be

6

Reposted by David Coeurjolly

Nicolas Bonneel @nbonneel.bsky.social · May 4

A thread on Monte Carlo integration, Sobol' sequences and our new awesome Siggraph paper "Sobol' Sequences with Guaranteed-Quality 2D Projections". by @dcoeurjo.bsky.social , J-C Iehl, V. Ostromoukhov and me. The tl;dr is our video: www.youtube.com/watch?v=8A4Z...
[1/16]

Sobol' Sequences with Guaranteed-Quality 2D Projections, supplementary video

YouTube video by Nicolas Bonneel

www.youtube.com

1 7 29

David Coeurjolly @dcoeurjo.bsky.social · May 5

Glad to share with you our new sampling paper just accepted at siggraph 2025 (@nbonneel.bsky.social , JC Iehl, V Ostromoukhov)

bsky.app/profile/nbon...

Nicolas Bonneel @nbonneel.bsky.social · May 4

A thread on Monte Carlo integration, Sobol' sequences and our new awesome Siggraph paper "Sobol' Sequences with Guaranteed-Quality 2D Projections". by @dcoeurjo.bsky.social , J-C Iehl, V. Ostromoukhov and me. The tl;dr is our video: www.youtube.com/watch?v=8A4Z...
[1/16]

Sobol' Sequences with Guaranteed-Quality 2D Projections, supplementary video

YouTube video by Nicolas Bonneel

www.youtube.com

5

Reposted by David Coeurjolly

Nicholas Sharp @nmwsharp.bsky.social · Apr 1

The Symposium on Geometry Processing is an amazing venue for geometry research: meshes, point clouds, neural fields, 3D ML, etc. Reviews are quick and high-quality.

The deadline is in ~10 days. Consider submitting your work, I'm planning to submit!

sgp2025.my.canva.site/submit-page-...

SGP 2025 - Submit page

sgp2025.my.canva.site