Accurate and Complete Surface Reconstruction from 3D Gaussians via Direct SDF Learning
This work addresses the problem of surface reconstruction for 3D scene modeling, offering a novel method that enhances geometric consistency in 3DGS, though it is incremental as it builds upon existing 3DGS techniques.
The paper tackles the challenge of achieving accurate and complete surface reconstruction from 3D Gaussian Splatting (3DGS) by proposing DiGS, a framework that integrates Signed Distance Field (SDF) learning directly into the pipeline, resulting in improved reconstruction accuracy and completeness on benchmarks like DTU, Mip-NeRF 360, and Tanks& Temples while maintaining high rendering fidelity.
3D Gaussian Splatting (3DGS) has recently emerged as a powerful paradigm for photorealistic view synthesis, representing scenes with spatially distributed Gaussian primitives. While highly effective for rendering, achieving accurate and complete surface reconstruction remains challenging due to the unstructured nature of the representation and the absence of explicit geometric supervision. In this work, we propose DiGS, a unified framework that embeds Signed Distance Field (SDF) learning directly into the 3DGS pipeline, thereby enforcing strong and interpretable surface priors. By associating each Gaussian with a learnable SDF value, DiGS explicitly aligns primitives with underlying geometry and improves cross-view consistency. To further ensure dense and coherent coverage, we design a geometry-guided grid growth strategy that adaptively distributes Gaussians along geometry-consistent regions under a multi-scale hierarchy. Extensive experiments on standard benchmarks, including DTU, Mip-NeRF 360, and Tanks& Temples, demonstrate that DiGS consistently improves reconstruction accuracy and completeness while retaining high rendering fidelity.