🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

NegGS: Negative Gaussian Splatting Artur Kasymov, Bartosz Czekaj, Marcin Mazur, Przemysław Spurek tl;dr: subtraction of probability density functions (PDFs) from two Gaussian distributions->density distribution->Gaussians with negative color arxiv.org/pdf/2405.18163

MiraGe: Editable 2D Images using Gaussian Splatting Contributions (cited): • We introduce the MiraGe model, which represents 2D images using flat Gaussian components, achieving state-of-the-art reconstruction quality. • MiraGe enables the manipulation of 2D images within 3D

In case you missed it, this is really cool. Image editing via Gaussian Splatting. We are getting here some fresh application ideas. It also comes with code to try it yourself.

Donald Tusk Donald Tusk Jan Grabiec IDEAS NCBR Stefan Dziembowski 🇵🇱 🇪🇺 🇺🇦 🏳️‍🌈 Przemyslaw Łukasz Kuciński Piotr Miłoś Marcin Moskalewicz Sylwia Czubkowska Maciej Kawecki - This Is IT anna wittenberg Bolesław Breczko Liderzy IDEAS NCBR dziękują Wam za wsparcie - razem możemy zmieniać świat na lepszy 🙌🚀💪 Thank you all for supporting IDEAS NCBR Piotr Sankowski and Polish science! Piotr Miłoś Leszek Bukowski 🧠💻🏛️👾 Michał Podlewski Adam Czarnecki Aʟᴇᴋꜱᴀɴᴅʀᴀ ᴘʀᴢᴇɢᴀʟɪɴꜱᴋᴀ Jaroslaw Krolewski Matthias Bethge and many more 🙌

If you are attending @ecai2024 and are interested in #machine_learning safety or Graph NNs please check out our work "Efficient Model-Stealing Attacks Against Inductive GNNs"! 📜 arxiv.org/abs/2405.12295 🗓️ October 21, 2024 ⏰ 12:15 PM – 12:30 PM 🎤 M87 talk by M. Podhajski

Neural Surface Priors for Editable Gaussian Splatting Jakub Szymkowiak, Weronika Jakubowska, Dawid Malarz, Weronika Smolak-Dyżewska, Maciej Zięba, Wojtek Pałubicki, Przemysław Musialski, Przemysław Spurek tl;dr: mesh-guided 3DGS editing arxiv.org/abs/2411.18311

⭐️ Another great success from our group! "FreSh: Frequency Shifting for Accelerated Neural Representation Learning" has been accepted for ICLR2025! Amazing work by Adam Kania, Marko Mihajlovic, Sergey Prokudin, Jacek Tabor and Przemyslaw Spurek!

🚀 Another great success from our team - Negative Gaussian Splatting (NegGS) is Accepted to Information Sciences! 📄🎉 Credits to Artur Kasymov, Bartosz Czekaj, Marcin Mazur, Jacek Tabor, and Przemysław Spurek for amazing work! sciencedirect.com/science/articl…

RaySplats: Ray Tracing based Gaussian Splatting Krzysztof Byrski, Marcin Mazur, Jacek Tabor, Tadeusz Dziarmaga, Marcin Kądziołka, Dawid Baran, Przemysław Spurek tl;dr: in title; distance between Gaussian ellipses and rays arxiv.org/abs/2501.19196

RaySplats: Ray Tracing based Gaussian Splatting Contributions: • We propose RaySplats, a novel differential rendering procedure for 3D Gaussian Splatting. It utilizes ray tracing during both the training and inference phases, yet does not require the use of polytope

Paper: arxiv.org/abs/2501.19196 Code: github.com/KByrski/RaySpl…

MeshSplats: Mesh-Based Rendering with Gaussian Splatting Initialization Rafał Tobiasz, Grzegorz Wilczyński, Marcin Mazur, Sławomir Tadeja, Przemysław Spurek tl;dr: flat Gaussians->disjoint meshes->ray tracing->mesh-based representation arxiv.org/abs/2502.07754

MeshSplats: Mesh-Based Rendering with Gaussian Splatting Initialization Contributions: • We introduce MeshSplats, a novel method facilitating the conversion of GS with flat Gaussians into a mesh-based representation, allowing further integration of dedicated tools. •

🚀 Code Release! 🚀 We just added code to our paper "β-CFG: Classifier-Free Guidance with Adaptive Scaling"! 🎉 🔹 Better image quality & prompt alignment 🔹 Adaptive normalization for stability 🔹 β-distribution scaling for optimization 💻 Code: github.com/gmum/beta-CFG GMUM - Group of Machine Learning Research

REdiSplats: Ray Tracing for Editable Gaussian Splatting Contributions: – REdiSplats enables handling lighting effects (such as reflections, shadows, transparency) and allows combining GS models with mesh-based models. – REdiSplats allows manual edition and physical simulation

Paper: arxiv.org/abs/2503.12284 Code: github.com/KByrski/REdiSp…

REdiSplats: Ray Tracing for Editable Gaussian Splatting Krzysztof Byrski, Grzegorz Wilczyński, Weronika Smolak-Dyżewska, Piotr Borycki, Dawid Baran, Sławomir Tadeja, Przemysław Spurek tl;dr: scene->flat Gaussian distributions parameterized by the mesh arxiv.org/abs/2503.12284

🚀 Excited to share an update on our latest work — RaySplats: Ray Tracing based Gaussian Splatting! We just added a viewer to our paper to let you explore the results interactively! 🎥✨ 🖱️ github.com/KByrski/RaySpl… #E3DGSAR #3DGaussianSplatting #FNP FNP Foundation for Polish Science GMUM - Group of Machine Learning Research

🚀 Excited to share our latest work: "CLIPGaussian: Universal and Multimodal Style Transfer Based on Gaussian Splatting"! GitHub: github.com/kornelhowil/CL… Project Page: kornelhowil.github.io/CLIPGaussian/ #GaussianSplatting #3DGS #FunduszeUE #FunduszeEuropejskie #E3DGSAR #FNP