RT-Splatting: Joint Reflection-Transmission Modeling with Gaussian Splatting

Hugging Face Daily Papers Papers
3d-gaussian-splatting reflection transmission novel-view-synthesis real-time-rendering scene-representation computer-graphics

Summary

RT-Splatting introduces a new 3D Gaussian Splatting framework that separates geometric occupancy from optical opacity to improve rendering of semi-transparent specular surfaces with high-fidelity reflections and transmission.

3D Gaussian Splatting (3DGS) enables real-time novel view synthesis with high visual quality. However, existing methods struggle with semi-transparent specular surfaces that exhibit both complex reflections and clear transmission, often producing blurry reflections or overly occluded transmission. To address this, we present RT-Splatting, a framework that disentangles each Gaussian's geometric occupancy from its optical opacity. This factorization yields a unified surface-volume scene representation with a single set of Gaussian primitives. Our hybrid renderer interprets this representation both as a surface to capture high-frequency reflections and as a volume to preserve clear transmission. To mitigate the ambiguity in jointly optimizing reflection and transmission, we introduce Specular-Aware Gradient Gating, which suppresses misleading gradients from highly specular regions into the transmission branch, effectively reducing distracting floaters. Experiments on challenging semi-transparent scenes show that RT-Splatting achieves state-of-the-art performance, delivering high-fidelity reflections and clear transmission with real-time rendering. Moreover, our factorization naturally enables flexible scene editing. The project page is available at https://sjj118.github.io/RT-Splatting.
Original Article
View Cached Full Text

Cached at: 05/20/26, 10:40 PM

Paper page - RT-Splatting: Joint Reflection-Transmission Modeling with Gaussian Splatting

Source: https://huggingface.co/papers/2605.18263 Published on May 18

·

Submitted byhttps://huggingface.co/sjj118

Ji Shion May 20

Abstract

RT-Splatting introduces a novel 3D Gaussian Splatting framework that separates geometric occupancy from optical opacity to improve rendering of semi-transparent specular surfaces with high-fidelity reflections and transmission.

3D Gaussian Splatting(3DGS) enables real-timenovel view synthesiswith high visual quality. However, existing methods struggle with semi-transparentspecular surfacesthat exhibit both complexreflections and cleartransmission, often producing blurryreflections or overly occludedtransmission. To address this, we present RT-Splatting, a framework that disentangles each Gaussian’sgeometric occupancyfrom itsoptical opacity. This factorization yields a unifiedsurface-volume scene representationwith a single set of Gaussian primitives. Ourhybrid rendererinterprets this representation both as a surface to capture high-frequencyreflections and as a volume to preserve cleartransmission. To mitigate the ambiguity in jointly optimizingreflectionandtransmission, we introduceSpecular-Aware Gradient Gating, which suppresses misleading gradients from highly specular regions into thetransmissionbranch, effectively reducing distractingfloaters. Experiments on challenging semi-transparent scenes show that RT-Splatting achieves state-of-the-art performance, delivering high-fidelityreflections and cleartransmissionwith real-time rendering. Moreover, our factorization naturally enables flexible scene editing. The project page is available at https://sjj118.github.io/RT-Splatting.

View arXiv pageView PDFProject pageGitHub55Add to collection

Get this paper in your agent:

hf papers read 2605\.18263

Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash

Models citing this paper0

No model linking this paper

Cite arxiv.org/abs/2605.18263 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2605.18263 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2605.18263 in a Space README.md to link it from this page.

Collections including this paper0

No Collection including this paper

Add this paper to acollectionto link it from this page.

Similar Articles

GlobalSplat: Efficient Feed-Forward 3D Gaussian Splatting via Global Scene Tokens

Hugging Face Daily Papers

GlobalSplat introduces an efficient feed-forward framework for 3D Gaussian splatting that achieves compact and consistent scene reconstruction using global scene tokens, reducing computational overhead and inference time to under 78ms. The method uses a coarse-to-fine training approach to prevent representation bloat while maintaining competitive novel-view synthesis performance with significantly fewer Gaussians (16K) compared to dense baselines.

Tsplat – Run Gaussian splatting in your terminal

Hacker News Top

Tsplat is a Rust-based CLI tool that renders 3D Gaussian Splatting scenes directly in the terminal using only the CPU, even over SSH. It supports Unicode half-blocks and graphics protocols, with controls for navigation and auto-orbit.

Gaussian Point Splatting

Hacker News Top

Researchers propose Gaussian Point Splatting, a stochastic rendering method using pixel-sized opaque points and 64-bit GPU atomics that scales to hundreds of millions of Gaussians in real time. The method, accepted at SIGGRAPH 2026, employs hierarchical culling and parallel programming primitives to achieve even workload distribution with only minor noise differences compared to original Gaussian splatting.