super cool & in-depth interactive intro/explainer about Physically Based Rendering imadr.me/pbr/ the author shares mostly on X ( > @imadr_ )

Well this was a delightful read. www.makingsoftware.com/chapters/col...

The Notebooks 2.0 gallery now has a dedicated page! observablehq.com/notebook-kit...

I wrote this article explaining algorithms for physarum-style simulations bleuje.com/physarum-exp...

GenAI FILP-FLOPS www.syntilay.com FINALLY!!

Full case study of Spirit Plant - our biggest project to date - is finally live! 🖼️🌱🌷 Learn more about - the installation concept - tech behind the generative plants and simulation - UI/UX design - and more... variable.io/expo-2025-sp...

another image to 3D model, excellent NV Synth, more faithful to the source image than TRELLIS, max 1536^3 resolution and it creates watertight meshes (=3D print ready)! lizhihao6.github.io/Sparc3D/ hf demo: huggingface.co/spaces/ilcve... not sure about the GPU requirements, probably high...

(credit @briankaris.bsky.social )

like MetaHuman but StyleGAN + Gaussian Splatting. it has the quirks of StyleGAN (asymmetrical / melty features) only renders half the model but it's precise (hairs & thin features) & provides a solid latent inversion and some tricks to improve the GS model. project: fraunhoferhhi.github.io/cgs-gan/

instant subscription :)

Today is the day! Inspect and tweak any @threejs.org website with ThreeTools Chrome extension. Registrations for the beta open at 5:00PM (CET) three.tools

After a hiatus, I am freelancing again ! available now for creative development work – In 2025 I would love to work on beautiful interfaces, storytelling & micro-interactions among other things :) 📮[email protected] 🛖 florianmorel.com (Remote from Taipei, UTC+8)

@froyok.fr this will sound like a bold move but since I can't use DMs, I'll give it a shot: hello, if you happen to live near Paris, I'd really like to meet you (professionally). I don't come back often to Paris and there's a "gathering" on the 19, 20 & 21st of may, this would be a good opportunity.

3D model generation often produces monolithic models with no semantics. here are some efforts in trying to make sense of the a 3D model to make them more usable (untested) PartField github.com/nv-tlabs/Par... SAMPart3D github.com/Pointcept/SA... Primitive Anything github.com/PrimitiveAny...

that's for when you need to estimate the pose of multiple pigeons in 3D => alexhang212.github.io/3D-MuPPET/

Please share widely: The EPFL Geometry Computing Laboratory is launching its third call for an Artist / Designer Residency for the fall of 2025! More information on how to apply here: www.epfl.ch/labs/gcm/ope... #design #art #residency #research

Not sure how long it has been around but it’s a really great ML resource huggingface.co/learn by @hf.co

very cool (live!) demo of what every major company is after these days : world emulation madebyoll.in/posts/world_...

trying out FramePack, I really like the "atlas" based approach, the exceptionally low GPU requirements, the consistency and how it handles the scene continuation in video generation lllyasviel.github.io/frame_pack_g... need to explore further but it's a keeper for now

working with genAI is literally like summoning a Mr. Meeseeks. from wikipedia: "[they] typically live for no more than a few hours in a constant state of pain [...] the longer an individual Meeseeks remains alive, the more insane and unhinged they become." en.wikipedia.org/wiki/Mr._Mee...

cool demo to test different strategies to find Nearest Neighbours latent.domains/midjourney-n... details on the methods in the about section (bottom left)

My new game is an open-ended programming toy about generating 3D objects from code. A free demo is available now on Steam! s.team/a/3401490

After 5 years in the making, I'm very happy to finally release Anime.js v4, my dream animation library: 🧩 New modular API ⚡️ Highly performant 🖱️ Scroll-linked animations 🫵 Draggables ↔️ Responsive ✨ Additive animations 🪽 WAAPI support And so much more! Check out the new website animejs.com!

the more I work with video generation (2 days...), the better I understand why - most of the time - generated videos show people posing and not doing shit. it's sufficient for advertising and pseudo trailers I guess but actually doing something with video generation takes a vicious amount of effort

can we vibe-code dependency hell yet?

apparently, Tencent made Hunyuan3D image to 3D ~30 times faster github.com/Tencent/Huny... (weights released 2 hours ago, need to try it)

super smart & very intuitive analogy to image edition, would never have thought of the "pathfinder" tool, well played! :)

stability's virtual camera looks interesting stability.ai/news/introdu...

the "Picture of Dorian Gray" rendered using Gemini-2's new inlined image generator (to test the consistency)

saw some pretty bad AI generated images on a "naturopathy" website today (it was recommended but muzli) and I think I'll never get used to people using generated images "as is", put them online and call it a day.

Dietails. All Braun, all from dasprogramm.co.uk

HI! You should watch JIMMY AND BABY! a 20 minute pixel art cartoon: youtu.be/zFbJ94jC1hc?...

TIL the CL1 is a thing corticallabs.com/cl1.html they coined the term 'Wetware' and provide 'Wetware-as-a-Service' which is as gross as it amazing. I do believe biological computers are the future (for many reasons) but it's the first time I see one

lovely collection

I heard using LLMs is called "vibe coding", not that it matters but I think that this name sucks and "rubber ducking" is more accurate

Sounds like a good time to look into these alternatives indeed 🙂

note to self, try this when back www.florafauna.ai

One month left to submit artwork made with/about computer vision to @cvprconference.bsky.social AI Art Gallery 😎 Details: bit.ly/CVPRAIArt25 Last year's gallery: thecvf-art.com #CVPR2025 #AIart

skeletons are tricky as fuck, wasted a day trying out various methods. most libs are not maintained or don't produce useful results, or only work on watertight meshes .... I ended up with a handmade voxel approach & have to simplify & vectorize the graph which is exactly what I wanted to avoid :/

progress on my TRELIIS to #Blender add-on, this is a "real-time" footage so it's not very exciting to watch but it gives a good idea of what it would be like to use #TRELLIS for real ✨

unsurprisingly Mv-DUSt3R ( github.com/facebookrese...) does not like 'turntable' footage and prefers continuous sequences. it can't replace - say - TRELLIS to convert few shots of an object into a 3D model. it's fast though, not sure how to prune discretize the result ( nice debug views too ^^)

TRELLIS works especially well on mechanical objects. in this case it made the mesh roughly symmetrical and fixed a missing leg. all the components are disjoint 'patches' scattered all over the mesh, disregarding the topology. I wonder how useful this can be for 3D artists, maybe as static assets?

to learn how to make an Blender add-on, I made a TRELLIS-Blender bridge. Blender is a powerful tool and not so horrible once you give up all hope and go with the (stack over)flow. wish I could explore further :) the add-on available here github.com/nicoptere/tr...

I managed to display a colored point cloud in Blender! it took longer than anticipated but it can be handy so I'll leave screenshots here. the import script is here pastebin.com/Dz2VusYH the PLY I used is a Gaussian Splat so it looks like shit and the R, G & B attributes are f_dc_0, f_dc_1 & f_dc_2

and that's one less in the "code coming soon" folder mv-dust3rp.github.io need to see how it fares with "turntable" footage to get object reconstruction instead of scene reconstruction (à la TRELLIS)

My game Stimulation Clicker is out now, enjoy :) > neal.fun/stimulation-...

Genesis (not the band) is a "looks good but what's the point & who's the audience exactly" kind of library. if the "LLM to 4D world" promise is dubious, the tech debt is very real and simply abysmal, see genesis-world.readthedocs.io/en/latest/us... & licenses / credits github.com/Genesis-Embo...

yet another "what could go wrong?" lib tangjiapeng.github.io/projects/GAF/

My kikk.be talk is finaly live. It's a wild ride through data art, brand alogorithms and custom design tools. I explain what is Variable Design and why it matters. www.youtube.com/watch?v=sBcr...

I'll be at @unireps.bsky.social this Saturday presenting a new experimental pipeline to visually explore structured neural network representations. The core idea is to take thousands of prompts that activate a concept, and then cluster and draw them using MultiDiffusion. 🧵👇