Aug 2026· IEEE Transactions on Visualization and Computer Graphics· Vol PP· 0 citations
Medicine
TL;DR
InfoLoD introduces a Fisher-guided self-distillation scheme that uses the Fisher Information Matrix to select geometrically valid, information-rich pseudo viewpoints, enabling LoD training directly from a pre-trained 3DGS model without any original images.
Abstract
Real-time visualization of massive 3D virtual environments, such as city-scale digital twins and immersive simulations, remains a core challenge in computer graphics. While 3D Gaussian Splatting (3DGS) achieves photorealistic rendering at interactive rates, its explicit Gaussian representation incurs substantial GPU memory overhead, limiting deployment on resource-constrained devices. Existing 3DGS Level-of-Detail (LoD) methods are often data-dependent, requiring access to large collections of original training images, and they rarely adapt to dynamic hardware budgets, leading to unstable performance on low-VRAM GPUs. We present InfoLoD, a data-free and hardware-aware framework for hierarchical LoD construction and large-scale 3DGS rendering. InfoLoD introduces a Fisher-guided self-distillation scheme that uses the Fisher Information Matrix (FIM) to select geometrically valid, information-rich pseudo viewpoints, enabling LoD training directly from a pre-trained 3DGS model without any original images. To operate under strict VRAM constraints, we further propose a budget-aware out-of-core streaming system with spatio-temporal working-set management and resource-adaptive LoD scheduling to balance fidelity and throughput. Experiments on challenging outdoor scenes show that InfoLoD enables stable real-time rendering on resource-constrained GPUs, delivering higher efficiency than prior baselines while maintaining competitive visual quality.
While 3D Gaussian Splatting (3DGS) has emerged as a powerful representation for real-time novel view synthesis, rendering high-fidelity scenes often relies on a massive number of Gaussian primitives, incurring substantial storage and computational overhead. Existing simplification techniques are largely intrusive, requiring training-time pruning, architectural modifications, or computationally expensive per-scene fine-tuning. These drawbacks limit their deployment on off-the-shelf pretrained models. In this paper, we propose CVT-GS, a novel optimization-free post-hoc simplification framework that directly compresses trained 3DGS scenes without sacrificing visual fidelity. Our approach first constructs spatially coherent cells over Gaussian centers via a geometry-aware Centroidal Voronoi Tessellation (CVT). Subsequently, a lightweight neural cell merger predicts the geometry and appearance of a single, highly representative Gaussian primitive for each cell under differentiable rendering supervision. By formulating simplification as a rendering-aware many-to-one merging process rather than naive primitive pruning, CVT-GS outputs a standard 3DGS scene that is seamlessly compatible with existing renderers. Experiments on various datasets demonstrate the superiority of our method. Notably, when achieving a 100-fold reduction in Gaussian points, our method operates 12 times faster than state-of-the-art methods while improving the PSNR by 1.3 dB.
Few-view surface reconstruction recovers the visible surfaces of a scene from a few posed RGB images, providing the 3D models that robots need to explore and interact online. On mobile platforms, the reconstruction must be fast and geometrically accurate while keeping a small memory footprint to ensure safe and efficient operation. 3D Gaussian Splatting (3DGS) offers a high-fidelity scene representation, but building it from a few views is ill-posed, as many distinct surfaces reproduce the same images, making traditional photometric methods prone to"floater"artifacts. End-to-end methods resolve the ambiguity by regressing splats with large, usually Transformer-based, networks that require heavy compute and memory while generalizing poorly to new scenes. We propose G2SR, which exploits a well-posed core of the task: given cross-view 2D splat correspondences, 3D splats follow analytically from multi-view geometry. G2SR employs a lightweight neural frontend to detect and track 2D Gaussian splats on the image plane and an analytic backend to triangulate each into a metric-scale 3D splat. On ScanNet, Replica, and DTU, G2SR matches or exceeds the geometric accuracy of state-of-the-art end-to-end methods while running at 69-89 reconstructions per second within 203 MB of GPU memory (5-107x less) for 2- and 3-view inputs at 384 x 512 resolution, offering a practical path to online Gaussian-based surface reconstruction.
Dasong Gao, Vivienne Sze, S. Karaman· arXiv.org· 0 citations
The rapid maturation of 3D Gaussian Splatting (3DGS) as a native output format of generative world models has created a compelling but underserved workflow gap: developers and artists who can prompt a photorealistic volumetric scene into existence lack the practical skills to deploy it as a navigable, interactive extended-reality (XR) experience. This 90-minute, hands-on technical workshop bridges that gap. Participants survey the landscape of spatial world models — contrasting video-based, point-cloud, and 3DGS representations across four levels of spatial information — then follow a structured five-stage pipeline: world model generation and scene export, WebXR integration with Spark.js 2.0, level-of-detail (LOD) tuning for real-time performance, semantic object detection via Splat Analyzer to surface interactive trigger zones, and open build time for creative extension. All tools are open-source or browser-accessible; the only prerequisite is Node.js 20 or higher. By the end of the session, every attendee has a live WebXR experience running on commodity XR hardware, informed by a conceptual map of where 3DGS world models sit in the broader ecosystem of generative spatial AI.
Nigel Joe Al Hartman, Alex Coulombe· Proceedings of the Special I...· 0 citations
We present SHARP-GS, a high-performance framework that unlocks real-time 8K rendering for immersive virtual experiences by overcoming the scaling bottlenecks of 3D Gaussian Splatting (3DGS). While 3DGS excels at real-time view synthesis, its performance degrades non-linearly at ultra-high resolutions due to excessive binning overhead, uncoalesced memory access, and redundant per-pixel arithmetic. SHARP-GS addresses these inefficiencies through three key contributions: (i) Resolution-Aware Adaptive Binning, which maintains a constant overlap factor via dynamic tile sizing and employs fine-grained sub-tile culling; (ii) a Morton-Ordered Memory Layout to ensure spatially coherent memory access; and (iii) Forward Differencing, which replaces expensive probability density function evaluations with efficient incremental updates. On high-end consumer GPUs, our framework achieves an average 2.55× speedup at 8K resolution, sustaining over 250 FPS with negligible visual quality loss. This performance delivers the necessary pixel throughput to enable high-fidelity, 120 Hz stereoscopic VR experiences. Code and data for this paper are at https://github.com/dingjr7/SHARP-GS.
Junran Ding, Weihang Liu, Yuke Li et al.· International Conference on...· 0 citations
Standard 3D Gaussian Splatting (3DGS) pipelines for Novel View Synthesis (NVS) are bottlenecked by Structurefrom-Motion (SfM) initialization. In casual, sparse-view scenarios, feature matching breaks down, causing the entire reconstruction process to fail. We replace this brittle dependency with a COLMAP-free, feed-forward initializer powered by a Visual Geometry Grounded Transformer (VGGT). By leveraging VGGT, our pipeline jointly estimates camera parameters and dense scene geometry across all views in a single pass. A Bridge Module then robustly normalizes the scene scale and conditions initial Gaussian opacity on geometric confidence to discourage floater artifacts during densification. Our framework reduces the initialization phase from minutes (full-scene SfM) to seconds and achieves $\mathbf{1 0 0} \boldsymbol{\%}$ initialization success from as few as three unposed images (a regime where COLMAP succeeds on only 1 of 7 Mip-NeRF 360 scenes). Project page: https://github.com/yuvanrajkrishna/VGGT-Sparse-3DGS.
Shreeya Venkatraman, Yuvan Raj Krishna, J. V.· International Conference on...· 0 citations
Free-viewpoint 3D scene media is increasingly important for immersive applications, yet practical capture often suffers from severe view sparsity and motion blur. Although neural rendering has advanced sparse-view synthesis, existing blur-aware methods typically require substantial multi-view redundancy, accurate camera poses, or costly per-scene optimization. We address a stringent yet practical setting: reconstructing a coherent 3D scene from only two motion-blurred images with known intrinsics, without input-view poses, auxiliary sharp images, or per-scene test-time optimization. To this end, we propose CasDeblurGS, a cascaded framework that progressively recovers reliable cross-view information from local 2D correspondences to global 3D guidance. Stage 1 constructs locally reliable guidance through occlusion-aware correspondence filtering, while Stage 2 aggregates the intermediate restorations into a provisional pose-free 3D Gaussian representation whose input-view re-renders provide dense global guidance for final restoration. The resulting views enable a more coherent 3D representation and higher-quality novel-view synthesis. Experiments on real-world and synthetic Deblur-NeRF scenes show consistent gains over strong baselines, improving PSNR by 1.19 dB and 2.11 dB, respectively. Progressive ablations, cross-view correspondence visualization, and camera reprojection analysis further demonstrate improvements in both rendering quality and multi-view geometric consistency.
Haeyun Choi, MinHyuk Jang, I-Gil Kim· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
