FreGS: 3D Gaussian Splatting with Progressive Frequency Regularization

TL;DR

FreGS introduces a progressive frequency regularization method for 3D Gaussian splatting that improves high-quality view synthesis by reducing over-reconstruction artifacts through frequency space optimization.

Contribution

It proposes a novel coarse-to-fine Gaussian densification technique using frequency space regularization to enhance 3D Gaussian splatting performance.

Findings

Outperforms state-of-the-art methods on multiple benchmarks.

Reduces over-reconstruction artifacts effectively.

Achieves superior novel view synthesis quality.

Abstract

3D Gaussian splatting has achieved very impressive performance in real-time novel view synthesis. However, it often suffers from over-reconstruction during Gaussian densification where high-variance image regions are covered by a few large Gaussians only, leading to blur and artifacts in the rendered images. We design a progressive frequency regularization (FreGS) technique to tackle the over-reconstruction issue within the frequency space. Specifically, FreGS performs coarse-to-fine Gaussian densification by exploiting low-to-high frequency components that can be easily extracted with low-pass and high-pass filters in the Fourier space. By minimizing the discrepancy between the frequency spectrum of the rendered image and the corresponding ground truth, it achieves high-quality Gaussian densification and alleviates the over-reconstruction of Gaussian splatting effectively. Experiments over multiple widely adopted benchmarks (e.g., Mip-NeRF360, Tanks-and-Temples and Deep Blending) show that FreGS achieves superior novel view synthesis and outperforms the state-of-the-art consistently.