EGGS: Exchangeable 2D/3D Gaussian Splatting for Geometry-Appearance Balanced Novel View Synthesis

TL;DR

EGGS introduces a hybrid 2D/3D Gaussian splatting method that balances appearance and geometric accuracy in novel view synthesis, outperforming existing approaches in quality and efficiency.

Contribution

The paper proposes Exchangeable Gaussian Splatting (EGGS), a novel hybrid representation combining 2D and 3D Gaussians with new techniques for unified rendering and dynamic adaptation.

Findings

EGGS achieves superior rendering quality compared to existing methods.

EGGS demonstrates improved geometric accuracy in novel view synthesis.

EGGS offers efficient training and inference through CUDA acceleration.

Abstract

Novel view synthesis (NVS) is crucial in computer vision and graphics, with wide applications in AR, VR, and autonomous driving. While 3D Gaussian Splatting (3DGS) enables real-time rendering with high appearance fidelity, it suffers from multi-view inconsistencies, limiting geometric accuracy. In contrast, 2D Gaussian Splatting (2DGS) enforces multi-view consistency but compromises texture details. To address these limitations, we propose Exchangeable Gaussian Splatting (EGGS), a hybrid representation that integrates 2D and 3D Gaussians to balance appearance and geometry. To achieve this, we introduce Hybrid Gaussian Rasterization for unified rendering, Adaptive Type Exchange for dynamic adaptation between 2D and 3D Gaussians, and Frequency-Decoupled Optimization that effectively exploits the strengths of each type of Gaussian representation. Our CUDA-accelerated implementation ensures efficient training and inference. Extensive experiments demonstrate that EGGS outperforms existing methods in rendering quality, geometric accuracy, and efficiency, providing a practical solution for high-quality NVS.