Javier Fabre

PhD student in computer graphics.
Render Software Engineer at @itsclo3d

[Javier Fabre @fjavifabre]

Our work "Fast Voxelization and LoD for Microgeometry Rendering" is now available in The Visual Computer!
https://t.co/IhTOeKTh5l

We introduce a fast parallel voxelization method and a novel SGGX LoD model to efficiently render complex microgeometry like fibers, trees or hair.

[Yining Karl Li @yiningkarlli]

There's a neat short paper from Eurographics this year that evaluates how well microfacet shading models actually predict measured reflectance. The answer is: not nearly as well as one would hope!

(1/3)

https://t.co/GcM5ZxYFvZ

[Javier Fabre @fjavifabre]

Project page: https://t.co/C7qpvedmCU

[Vlad Erium 🇯🇵 @ssh4net]

Fast Voxelization and Level of Detail for Microgeometry Rendering

Javier Fabre, Carlos Castillo, Carlos Rodriguez-Pardo, Jorge Lopez-Moreno

Paper: https://t.co/LfIe3egpxA
Supplementary: https://t.co/c7f3QyROdl

Abstract:
Many materials show anisotropic light scattering patterns due to the shape and local alignment of their underlying micro structures: surfaces with small elements such as fibers, or the ridges of a brushed metal, are very sparse and require a high spatial resolution to be properly represented as a volume. The acquisition of voxel data from such objects is a time and memory-intensive task, and most rendering approaches require an additional Level-of-Detail (LoD) data structure to aggregate the visual appearance, as observed from multiple distances, in order to reduce the number of samples computed per pixel (E.g.: MIP mapping). In this work we introduce first, an efficient parallel voxelization method designed to facilitate fast data aggregation at multiple resolution levels, and second, a novel representation based on hierarchical SGGX clustering that provides better accuracy than baseline methods. We validate our approach with a CUDA-based implementation of the voxelizer, tested both on triangle meshes and volumetric fabrics modeled with explicit fibers. Finally, we show the results generated with a path tracer based on the proposed LoD rendering model.

[Markus Schütz @m_schuetz]

New Paper🙂

Nanite has shown that small triangles can be rendered fast in compute, we're exploring how fast for large meshes with up to 18.9 billion triangles, without the need to precompute LOD structures.

Paper: https://t.co/F9u4xE6Na3

Source: https://t.co/1LgdSHVi7i

[Vlad Erium 🇯🇵 @ssh4net]

Fast Voxelization and Level of Detail for Microgeometry Rendering

Javier Fabre, Carlos Castillo, Carlos Rodriguez-Pardo, Jorge Lopez-Moreno

Paper: https://t.co/LfIe3egpxA
Supplementary: https://t.co/c7f3QyROdl

Abstract:
Many materials show anisotropic light scattering patterns due to the shape and local alignment of their underlying micro structures: surfaces with small elements such as fibers, or the ridges of a brushed metal, are very sparse and require a high spatial resolution to be properly represented as a volume. The acquisition of voxel data from such objects is a time and memory-intensive task, and most rendering approaches require an additional Level-of-Detail (LoD) data structure to aggregate the visual appearance, as observed from multiple distances, in order to reduce the number of samples computed per pixel (E.g.: MIP mapping). In this work we introduce first, an efficient parallel voxelization method designed to facilitate fast data aggregation at multiple resolution levels, and second, a novel representation based on hierarchical SGGX clustering that provides better accuracy than baseline methods. We validate our approach with a CUDA-based implementation of the voxelizer, tested both on triangle meshes and volumetric fabrics modeled with explicit fibers. Finally, we show the results generated with a path tracer based on the proposed LoD rendering model.

[solst/ICE of Astarte @IceSolst]

Holy shit they’ve LinkedInified my shitpost, including a medium article, and tons of ai hallucinated corpospeak, truly a cursed platform

[Yining Karl Li @yiningkarlli]

So I have a 200MB FBX file, and it takes >6 hours to open in Maya 2024 and <10 seconds to open in Blender 4.2.3 LTS. ¯\_(ツ)_/¯

[Jorge Condor @Arcanous98]

Heads up to everyone, I'll be in Vancouver (from Saturday) to present our paper on primitive-based, physically-based volume rendering at SIGGRAPH!

If anyone wants to chat feel free to drop a DM :D

Link to the paper: https://t.co/4AKZ5vKPOi

[Markus Schütz @m_schuetz]

Just tried OpenGL bindless and it's beautiful 🥹

- Easily drawing all scene nodes with geometry in separate storage buffers in a single draw call.
- Vertex shader gets a list of nodes, each node contains pointer to the node's geometry.

No vertex buffers, no bindings. 🥹

[Christian Richardt @c_richardt]

I don’t understand why someone would publicly lie about being a coauthor on a paper when it’s so easy to prove wrong?!

Shame on @ChuanRichardLi.

[Darren Parkinson 🌍💚🤘 @DarrenBar88]

Oh, and if you haven't already, you can watch the short film that the Guardian made about my life living with #LongCovid here ⬇️

https://t.co/4bjiNvIkyJ

[Valentin Deschaintre @vdeschaintre]

🎓 We introduce SAMa! A material selection and segmentation model on 3D models in any format (3DGS, NeRF, Mesh).
Given a user click, we propose to select all regions on an objects with the same material. We can also do segmentation in under a minute: https://t.co/cl3KvzBLcw

[Vlad Erium 🇯🇵 @ssh4net]

Blendify -- Python rendering framework for Blender
Vladimir Guzov, Ilya A. Petrov, Gerard Pons-Moll

Paper: https://t.co/dEj4W34FDk
Code (GPL3.0): https://t.co/3xhswjwvQu

Abstract:
With the rapid growth of the volume of research fields like computer vision and computer graphics, researchers require effective and user-friendly rendering tools to visualize results. While advanced tools like Blender offer powerful capabilities, they also require a significant effort to master. This technical report introduces Blendify, a lightweight Python-based framework that seamlessly integrates with Blender, providing a high-level API for scene creation and rendering. Blendify reduces the complexity of working with Blender's native API by automating object creation, handling the colors and material linking, and implementing features such as shadow-catcher objects while maintaining support for high-quality ray-tracing rendering output. With a focus on usability Blendify enables efficient and flexible rendering workflow for rendering in common computer vision and computer graphics use cases.

#cv #blender #b3d #python

[Dan Casas @dancasas]

New curated SMPL texture batch just dropped! Now including sports outfits 🏀🏈.

Get them for free at https://t.co/Pja0tF9Do6 or sample SMPLitex to generate new ones!

[Javier Fabre @fjavifabre]

Really cool!

I remember programming 10+ times some sampling techniques in a quick python program just to view the way samples look in a 2D / 3D plot.

Plus it runs in your browser!!

[Wojciech Jarosz @wkjarosz]

I'm calling it. Samplin' Safari version 2 is out! This is my small research-oriented tool to visualize and interactively inspect high-dimensional random points (as used for e.g. Monte Carlo #rendering).

Many changes and new features (a thread).

https://t.co/LSKedXNFoh

[Diego Gutierrez @diego__guti]

Pregunta a los investigadores de las universidades españolas: Meta quiere investigar temas de VR conmigo, para lo cual hay que firmar un NDA bastante corriente. El problema es que mi universidad quiere meter una cláusula con la que no estoy de acuerdo: 👇

[Diego Gutierrez @diego__guti]

Our work to be presented at #sigAsia2023: a fully-differentiable end-to-end NLOS inverse rendering pipeline that self-calibrates the imaging parameters! We optimally reconstruct geometry and albedo of hidden scenes. Joint work with KAIST and @GILabUZ @julio_marco

[Dan Casas @dancasas]

Yesterday, I had the honor of receiving the medal from the Royal Academy of Engineering of Spain (@RAI_Ingenieria) as a runner-up for the Young Researcher Award 2023.

Thanks students and collaborators for the great work, and family for all the support!

[Javier Fabre @fjavifabre]

@dancasas @RAI_Ingenieria Enhorabuena!!

[Bart Wronski 🇺🇦🇵🇸 @BartWronsk]

I3D 2024 ACM Symposium on Interactive 3D Graphics and Games Call for Papers is open!
https://t.co/g5xGd8mZ0i
Consider submitting any innovative 3D work - academic research or the industry.
I am the conference's Paper Co-Chair this year, so if you have any questions - contact me!