In Depth We Trust: Reliable Monocular Depth Supervision for Gaussian Splatting

TL;DR

This paper presents a training framework that effectively incorporates monocular depth priors into Gaussian Splatting, improving 3D rendering quality despite depth estimation ambiguities.

Contribution

It introduces a novel method to handle scale ambiguity and noise in monocular depth maps, enhancing geometric accuracy in Gaussian Splatting.

Findings

Consistent improvements in geometric accuracy across datasets.

Enhanced rendering quality with monocular depth priors.

Effective isolation of ill-posed geometry for regularization.

Abstract

Using accurate depth priors in 3D Gaussian Splatting helps mitigate artifacts caused by sparse training data and textureless surfaces. However, acquiring accurate depth maps requires specialized acquisition systems. Foundation monocular depth estimation models offer a cost-effective alternative, but they suffer from scale ambiguity, multi-view inconsistency, and local geometric inaccuracies, which can degrade rendering performance when applied naively. This paper addresses the challenge of reliably leveraging monocular depth priors for Gaussian Splatting (GS) rendering enhancement. To this end, we introduce a training framework integrating scale-ambiguous and noisy depth priors into geometric supervision. We highlight the importance of learning from weakly aligned depth variations. We introduce a method to isolate ill-posed geometry for selective monocular depth regularization, restricting the propagation of depth inaccuracies into well-reconstructed 3D structures. Extensive experiments across diverse datasets show consistent improvements in geometric accuracy, leading to more faithful depth estimation and higher rendering quality across different GS variants and monocular depth backbones tested.