iiinova — where your digital self keeps evolving

Reposted by Florian Hahlbohm

Josip Krapac @josipk.bsky.social · 15/03/2026

But if your PhD took place in an environment that nurtured genuine curiosity, where you drove your own research, defended your ideas through debate, and developed the ability to ask meaningful questions, then the notion of LLMs conducting research probably doesn't sit right with you.

072

Reposted by Florian Hahlbohm

SIGBOVIK @harryqbovik.bsky.social · 27/02/2026

SIGBOVIK has a Bluesky now! Follow to learn more cutting-edge research from the world’s most comedic and occasionally scientific academic conference

06525

Reposted by Florian Hahlbohm

Thomas Wimmer @wimmerthomas.bsky.social · 28/03/2025

Had the honor to present "Gaussians-to-Life" at #3DV2025 yesterday. In this work, we used video diffusion models to animate arbitrary 3D Gaussian Splatting scenes. This work was a great collaboration with @moechsle.bsky.social, @miniemeyer.bsky.social, and Federico Tombari. 🧵⬇️

2131

Reposted by Florian Hahlbohm

Andreea Ardelean @andreead-a.bsky.social · 26/03/2025

Had a great experience presenting our work on 3D scene reconstruction from a single image with @visionbernie.bsky.social at #3DV2025 🇸🇬 andreeadogaru.github.io/Gen3DSR Reach out if you're interested in discussing our research or exploring international postdoc opportunities @fau.de

0184

Reposted by Florian Hahlbohm

Markus Schütz @m-schuetz.bsky.social · 21/03/2025

Here is our gaussian splat editor: github.com/m-schuetz/Sp... Eventually I want it to be able to take scans of ugly streets and beautify them like a photoshop for gaussians. :)

0143

Florian Hahlbohm @fhahlbohm.bsky.social · 10/03/2025

"DaD's a pretty good keypoint detector, probably the best." Nice one 😂

140

Florian Hahlbohm @fhahlbohm.bsky.social · 08/03/2025

We also provide a multitude of data loaders, camera model implementations, as well as various utilities for optimization and visualization.

020

Florian Hahlbohm @fhahlbohm.bsky.social · 08/03/2025

Each method has a Trainer, Model, and Renderer class that extend the respective base classes. Many of the current methods also define custom CUDA extensions or a designated loss class.

110

Florian Hahlbohm @fhahlbohm.bsky.social · 08/03/2025

NeRFICG is a research-focused framework for developing novel view synthesis methods. Shoutout to my colleague Moritz Kappel, who is responsible for most of the underlying architecture! We think, NeRFICG is a decent starting point for any PyTorch-based graphics/vision project.

110

Florian Hahlbohm @fhahlbohm.bsky.social · 08/03/2025

Further discussion and ideas for where things could be improved can be found in our paper and the "Additional Notes" in our GitHub repository. The remainder is on our framework NeRFICG: github.com/nerficg-proj...

github.com

NeRFICG

A flexible Pytorch framework for simple and efficient implementation of neural radiance fields and rasterization-based view synthesis methods. - NeRFICG

110

Florian Hahlbohm @fhahlbohm.bsky.social · 08/03/2025

BlueSky did not let me have two videos in the same post. So here's the OIT video.

110

Florian Hahlbohm @fhahlbohm.bsky.social · 08/03/2025

An interesting observation we had is that OIT (enabled by setting "Blend Mode" to 3 in the config) seems to help background reconstruction and overall densification. Videos show the first 3K training iterations using hybrid vs. order-independent transparency.

110

Florian Hahlbohm @fhahlbohm.bsky.social · 08/03/2025

Note that the GUI has a non-negligible impact on frame rate as it is Python-based. So you won't see maximum performance even after turning off v-sync. It is also Linux-only but my colleague Timon Scholz recently started working on a C++ version that also supports Windows.

120

Florian Hahlbohm @fhahlbohm.bsky.social · 08/03/2025

Btw, all visualizations in this thread use our perspective-correct approach for rendering 3D Gaussians. It is based on ray-casting and can be implemented efficiently. However the high frame rates reported in our paper are due to the hybrid transparency approach.

110

Florian Hahlbohm @fhahlbohm.bsky.social · 08/03/2025

Here are examples using (0) hybrid transparency with K=16, (1) alpha blending of the 4 first fragments per-pixel, (2) alpha blending in "global" depth-ordering, and (3) order-independent transparency. Model was trained using the settings as in (0).

120

Florian Hahlbohm @fhahlbohm.bsky.social · 08/03/2025

You can also modify the "Blend Mode" (see the Readme on GitHub) and core size K for blending modes where this is applicable. To reduce compile times, we only compile kernels for K in [1, 2, 4, 8, 16, 32] and "round down" for other values (e.g., 12 -> 8).

120

Florian Hahlbohm @fhahlbohm.bsky.social · 08/03/2025

Via the "Viewer Config" (F3), you can switch to rendering depth maps and expanding the advanced renderer config allows you to switch between expected (shown here) and median depth.

120

Florian Hahlbohm @fhahlbohm.bsky.social · 08/03/2025

Don't get confused by the "Time" stuff, which is for dynamic scenes reconstructed by methods such as our recent D-NPC: github.com/MoritzKappel... HTGS also does currently not support changing the background color and using camera models other than "Perspective" without distortion.

github.com

GitHub - MoritzKappel/D-NPC: Official code release for "D-NPC: Dynamic Neural Point Clouds for Non-Rigid View Synthesis from Monocular Video".

Official code release for "D-NPC: Dynamic Neural Point Clouds for Non-Rigid View Synthesis from Monocular Video". - MoritzKappel/D-NPC

120

Florian Hahlbohm @fhahlbohm.bsky.social · 08/03/2025

By modifying the "Principal Point" and/or "Focal Length" you can create fun images like the one below. You can even do this while watching your Gaussians train if you set TRAINING.GUI.ACTIVATE to true in the config file. And yes, you could in theory train on images like this.

120

Florian Hahlbohm @fhahlbohm.bsky.social · 08/03/2025

Let's start with the GUI features you might want to try with HTGS. If you open the "Camera Config" panel (F4) you can switch between "Orbital" and "Walking" controls. You can also modify the near/far plane.

120

Florian Hahlbohm @fhahlbohm.bsky.social · 08/03/2025

Many thanks to my co-authors Fabian Friederichs, @timweyrich.bsky.social, @linusfranke.bsky.social, Moritz Kappel, Susana Castillo and @mcstammi.bsky.social, Martin Eisemann, and Marcus Magnor! Thoughts and things to try in the thread below:

131

Florian Hahlbohm @fhahlbohm.bsky.social · 08/03/2025

We recently released the code for "Efficient Perspective-Correct 3D Gaussian Splatting Using Hybrid Transparency" Project Page: fhahlbohm.github.io/htgs/ Code: github.com/nerficg-proj...

2208

Reposted by Florian Hahlbohm

Arthur @arthurperpixel.bsky.social · 30/01/2025

I’ve released a new version of my 3D reconstruction tool, Brush 🖌️ It's a big step forward - the quality & speed now match gsplat, and there’s a lot of other new features! See the release notes github.com/ArthurBrusse... Some of the new features:

A autumnal stump, covered in mushrooms. This is a still from the interactive 3D reconstruction!

2278

Reposted by Florian Hahlbohm

Dmytro Mishkin @ducha-aiki.bsky.social · 23/01/2025

@chrisoffner3d.bsky.social we are in need of your eval)

051

Florian Hahlbohm @fhahlbohm.bsky.social · 25/12/2024

Merry Christmas :) I tried this as well but with Brush by @arthurperpixel.bsky.social . How many pictures did you take? For me Colmap only ended up using like 25/50 images and it didn't work that well. Tbf lighting was pretty bad.

100

Reposted by Florian Hahlbohm

Zhenjun Zhao @ericzzj.bsky.social · 05/12/2024

Volumetrically Consistent 3D Gaussian Rasterization Chinmay Talegaonkar, Yash Belhe, Ravi Ramamoorthi, Nicholas Antipa tl;dr: volumetrically integrate 3D Gaussians directly to compute the transmittance across them analytically->physically-accurate alpha values arxiv.org/abs/2412.03378

041

Florian Hahlbohm @fhahlbohm.bsky.social · 01/12/2024

I really enjoyed watching the videos the last time you did this. Thanks for making them available to everyone :)

020