VEGS: View Extrapolation of Urban Scenes in 3D Gaussian Splatting using Learned Priors

TL;DR

This paper introduces VEGS, a novel approach for extrapolated view synthesis in urban scenes using 3D Gaussian splatting, dense LiDAR initialization, and learned priors to enhance rendering quality outside training camera views.

Contribution

We propose the first method addressing Extrapolated View Synthesis in urban scene reconstruction, combining LiDAR-based initialization and learned scene priors for improved rendering.

Findings

Effective EVS performance demonstrated through qualitative results.

Quantitative improvements over baseline methods.

First to address EVS in urban scene reconstruction.

Abstract

Neural rendering-based urban scene reconstruction methods commonly rely on images collected from driving vehicles with cameras facing and moving forward. Although these methods can successfully synthesize from views similar to training camera trajectory, directing the novel view outside the training camera distribution does not guarantee on-par performance. In this paper, we tackle the Extrapolated View Synthesis (EVS) problem by evaluating the reconstructions on views such as looking left, right or downwards with respect to training camera distributions. To improve rendering quality for EVS, we initialize our model by constructing dense LiDAR map, and propose to leverage prior scene knowledge such as surface normal estimator and large-scale diffusion model. Qualitative and quantitative comparisons demonstrate the effectiveness of our methods on EVS. To the best of our knowledge, we are the first to address the EVS problem in urban scene reconstruction. Link to our project page: https://vegs3d.github.io/.