arXiv:2511.17364v1 Announce Type: new 
Abstract: We extend the recently proposed sparse voxel rasterization paradigm to the task of high-fidelity surface reconstruction by integrating Signed Distance Function (SDF), named SVRecon. Unlike 3D Gaussians, sparse voxels are spatially disentangled from their neighbors and have sharp boundaries, which makes them prone to local minima during optimization. Although SDF values provide a naturally smooth and continuous geometric field, preserving this smoothness across independently parameterized sparse voxels is nontrivial. To address this challenge, we promote coherent and smooth voxel-wise structure through (1) robust geometric initialization using a visual geometry model and (2) a spatial smoothness loss that enforces coherent relationships across parent-child and sibling voxel groups. Extensive experiments across various benchmarks show that our method achieves strong reconstruction accuracy while having consistently speedy convergence. The code will be made public.

SVRecon هي طريقة جديدة تعزز من تقنية التراص النقطي المتناثر لإعادة بناء السطوح بدقة عالية من خلال دمج دالة المسافة الموقعة (SDF). تتناول هذه الطريقة التحديات المتعلقة بالحد الأدنى المحلي أثناء التحسين وتهدف إلى الحفاظ على السلاسة عبر الفوكسلات المتناثرة المهيكلة من خلال بدء هندسي قوي وفقدان سلاسة مكاني.

SVRecon es un nuevo método que mejora la rasterización de vóxeles dispersos para la reconstrucción de superficies de alta fidelidad al integrar la Función de Distancia Firmada (SDF). Este enfoque aborda los desafíos relacionados con los mínimos locales durante la optimización y busca mantener la suavidad entre los vóxeles dispersos parametrizados mediante una inicialización geométrica robusta y una pérdida de suavidad espacial.

SVRecon est une nouvelle méthode qui améliore la rasterisation de voxels clairsemés pour la reconstruction de surfaces de haute fidélité en intégrant la fonction de distance signée (SDF). Cette approche aborde les défis liés aux minima locaux lors de l'optimisation et vise à maintenir la douceur entre les voxels clairsemés paramétrés grâce à une initialisation géométrique robuste et à une perte de douceur spatiale.

SVRecon is a newly proposed method that enhances sparse voxel rasterization for high-fidelity surface reconstruction by integrating Signed Distance Function (SDF). This approach addresses challenges related to local minima during optimization and aims to maintain smoothness across parameterized sparse voxels through robust geometric initialization and a spatial smoothness loss.

SVRecon: Sparse Voxel Rasterization for Surface Reconstruction

arXiv:2512.07107v2 Announce Type: replace 
Abstract: We present COREA, the first unified framework that jointly learns relightable 3D Gaussians and a Signed Distance Field (SDF) for accurate geometry reconstruction and faithful relighting. While recent 3D Gaussian Splatting (3DGS) methods have extended toward mesh reconstruction and physically-based rendering (PBR), their geometry is still learned from 2D renderings, leading to coarse surfaces and unreliable BRDF-lighting decomposition. To address these limitations, COREA introduces a coarse-to-fine bidirectional 3D-to-3D alignment strategy that allows geometric signals to be learned directly in 3D space. Within this strategy, depth provides coarse alignment between the two representations, while depth gradients and normals refine fine-scale structure, and the resulting geometry supports stable BRDF-lighting decomposition. A density-control mechanism further stabilizes Gaussian growth, balancing geometric fidelity with memory efficiency. Experiments on standard benchmarks demonstrate that COREA achieves superior performance in novel-view synthesis, mesh reconstruction, and PBR within a unified framework.

تم تقديم COREA كإطار رائد يجمع بين الغاوسيات ثلاثية الأبعاد القابلة لإعادة الإضاءة وحقول المسافة الموقعة (SDF) لتحسين إعادة بناء الهندسة ودقة إعادة الإضاءة. تستخدم هذه الطريقة استراتيجية محاذاة ثنائية الاتجاه من الخشن إلى الدقيق، مما يسمح بتحسين تعلم الإشارات الهندسية مباشرة في الفضاء ثلاثي الأبعاد، مما يعالج القيود التي لوحظت في الطرق السابقة لتوزيع الغاوسيات ثلاثية الأبعاد.

COREA se ha presentado como un marco pionero que integra Gaussianos 3D relightables y Campos de Distancia Firmados (SDF) para mejorar la reconstrucción geométrica y la precisión del relighting. Este enfoque emplea una estrategia de alineación bidireccional de grueso a fino, permitiendo un mejor aprendizaje de señales geométricas directamente en el espacio 3D, abordando las limitaciones observadas en métodos anteriores de Splatting Gaussiano 3D.

COREA a été introduit comme un cadre pionnier qui intègre des Gaussiens 3D relightables et des Champs de Distance Signés (SDF) pour améliorer la reconstruction géométrique et la précision du relighting. Cette approche utilise une stratégie d'alignement bidirectionnel de gros à fin, permettant un meilleur apprentissage des signaux géométriques directement dans l'espace 3D, répondant aux limitations observées dans les méthodes précédentes de Splatting Gaussien 3D.

COREA has been introduced as a pioneering framework that integrates relightable 3D Gaussians and Signed Distance Fields (SDF) to enhance geometry reconstruction and relighting accuracy. This approach employs a coarse-to-fine bidirectional alignment strategy, allowing for improved geometric signal learning directly in 3D space, addressing limitations seen in previous 3D Gaussian Splatting methods.

SVRecon: Sparse Voxel Rasterization for Surface Reconstruction

Was this article worth reading? Share it

LucidQuery AI

Deptho.ai

VSDECO