Developing more precise per-Gaussian normal estimation

Develop more precise per-Gaussian surface normal estimation techniques for ray-tracing-based Gaussian splatting, including HybridSplat, to improve reflection modeling fidelity and mitigate the current need for better normals highlighted by the authors.

Background

Accurate surface normals are critical for reflection modeling in ray-tracing-based Gaussian splatting, as normals determine reflection ray directions and the blending of reflective contributions. The HybridSplat framework relies on normals of reflective Gaussians; the authors explicitly state that more precise normal estimation is still needed, aligning with similar needs in related methods.

This explicit need is marked as future work by the authors, making it a clear open problem rather than a general limitation or aspiration.

References

One main drawbacks of our reflection-baked Gaussian tracing comes from the very near-field reflections in single object scenario, and also more precise Gaussian normal estimation are still needed as like other ray-tracing Gaussian splatting methods, which we leave for future works.

— HybridSplat: Fast Reflection-baked Gaussian Tracing using Hybrid Splatting  (2512.08334 - Liu et al., 9 Dec 2025) in Conclusion (Section 6)

Such scenes would require detecting multiple distinct specular reflections to compute several surface normals. This introduces a correspondence problem, as it must be determined which specular reflection originates from which light source. Thus, a more complex reconstruction strategy is required, and hence remains a subject for future work.

— Disparity Estimation of Planar Reflective Surfaces Using Specular Reflections From a Single Light Source  (2609.24756 - Kossira et al., 21 Sep 2026) in Section 5.4.3, “Illumination Boundaries”