SplashSplat introduces a new benchmark of 20 real-world splashing liquid scenes captured with seven synchronized 4K cameras at 60 fps, providing per-view liquid and container masks and fixed evaluation splits. The method reconstructs per‑frame liquid signed distance fields (SDFs) from these masks, fuses them into a coarse velocity field, and uses Lagrangian carriers to generate differentiable local Gaussian representations for rendering. SplashSplat outperforms existing dynamic Gaussian splatting techniques on both real and synthetic data, offering more physically plausible motion, lower training cost, and enabling temporal interpolation and style transfer without re‑optimization.
arXiv:2607. 25321v1 Announce Type: new Abstract: Video diffusion models generate visually compelling content but routinely violate elementary physics when the subject involves fluids: liquid columns break apart in mid-air, container water levels fail to rise as liquid is poured in, and splashes disperse without regard to momentum or gravity.
The paper introduces a technique for animating water within static 3D Gaussian Splatting (3DGS) scenes, allowing realistic, looping water motion without dynamic capture or physical simulation. It estimates 2D optical flow from images or renderings, lifts it into a 3D Eulerian motion field, and deforms Gaussian primitives accordingly, producing temporally seamless animations. The method maintains real‑time rendering performance and improves visual quality and efficiency compared to prior dynamic Gaussian splatting and 3D cinemagraph approaches.
The paper introduces a technique for animating water within static 3D Gaussian Splatting (3DGS) scenes, allowing realistic motion without dynamic capture or physical simulation. It estimates 2D optical flow from images or renderings, lifts it into a 3D Eulerian motion field, and applies this to deform Gaussian primitives, creating temporally seamless looping animations. The method maintains real‑time rendering performance, improves perceptual realism, and supports free‑viewpoint navigation, outperforming prior dynamic Gaussian splatting and 3D cinemagraph approaches on a new waterfall dataset.
arXiv:2609.38488v1 Announce Type: cross Abstract: Conventional 3D Gaussian Splatting (3DGS) requires depth sorting and ordered alpha blending to correctly render overlapping Gaussian primitives. Stoc...
arXiv:2610.28257v1 Announce Type: new Abstract: This work presents SplatFlux, a feed-forward latent-to-4DGS framework for both video-based 4D reconstruction and conditional 4D generation. Our key obs...
arXiv:2609.36940v1 Announce Type: new Abstract: Accurate dynamic scene reconstruction is important for robotic perception, where temporally consistent representations of dynamic environments are esse...
arXiv:2608.22906v1 Announce Type: new Abstract: Recent monocular 3D Gaussian Splatting (3DGS) streaming reconstruction methods have achieved impressive performance by balancing reconstruction quality...
arXiv:2608.21828v1 Announce Type: new Abstract: Novel-view synthesis of dynamic scenes, crucial for AR/VR applications, remains a challenging problem. Recent methods adapt representations like 3D Gau...
ResStreamGS introduces a residual‑sensitive dual‑stream framework for mobile volumetric video streaming that distinguishes temporally stable Gaussians from residual‑sensitive ones. Stable Gaussians are transmitted once per Group of Pictures, while only the residual‑sensitive Gaussians are included in the frame‑wise video stream, reducing bandwidth usage. Experiments on complex full‑scene videos show a 7.51–8.35 dB PSNR improvement over a V³‑style baseline, a payload reduction of up to 56.5%, and real‑time performance of 30.883 FPS on a Qualcomm Snapdragon 778G.
arXiv:2601.15951v2 Announce Type: replace Abstract: Novel view synthesis (NVS) of static and dynamic urban scenes is essential for autonomous driving simulation, yet existing methods often struggle t...
Recent monocular 3D Gaussian Splatting (3DGS) streaming reconstruction methods have achieved impressive performance by balancing reconstruction quality and efficiency. However, extending these framewo...
NoPo4D is a feed‑forward system that reconstructs dynamic 3D scenes from unposed multi‑view videos by using 4D Gaussian splatting. It introduces a velocity decomposition that separates per‑pixel image‑plane shifts from depth changes, enabling supervision with pseudo optical flow and avoiding the need for accurate camera poses or 3D motion ground truth. The method also employs a bidirectional motion encoder and view‑dependent opacity to improve cross‑view and cross‑timestep alignment, achieving state‑of‑the‑art results on multiple benchmarks while being much faster than per‑scene optimization approaches.