Neural Shell Texture Splatting: More Details and Fewer Primitives

TL;DR

This paper introduces neural shell texture, a global representation that disentangles geometry and appearance in Gaussian splatting, enabling high-quality, detailed textures with fewer primitives for view synthesis.

Contribution

It proposes a neural shell texture method that improves parameter efficiency and texture detail in Gaussian splatting by disentangling geometry and appearance.

Findings

Achieves high-fidelity view synthesis with fewer primitives.

Enables fine texture detail reconstruction.

Facilitates easy textured mesh extraction.

Abstract

Gaussian splatting techniques have shown promising results in novel view synthesis, achieving high fidelity and efficiency. However, their high reconstruction quality comes at the cost of requiring a large number of primitives. We identify this issue as stemming from the entanglement of geometry and appearance in Gaussian Splatting. To address this, we introduce a neural shell texture, a global representation that encodes texture information around the surface. We use Gaussian primitives as both a geometric representation and texture field samplers, efficiently splatting texture features into image space. Our evaluation demonstrates that this disentanglement enables high parameter efficiency, fine texture detail reconstruction, and easy textured mesh extraction, all while using significantly fewer primitives.