GGS: Generalizable Gaussian Splatting for Lane Switching in Autonomous Driving

TL;DR

GGS introduces a novel Gaussian Splatting approach for autonomous driving that enables realistic lane switching under large viewpoint changes, even with limited data, by generating virtual lanes and refining depth estimation.

Contribution

The paper presents a new virtual lane generation module, a diffusion loss for virtual lane supervision, and a depth refinement technique, advancing large viewpoint rendering in autonomous driving.

Findings

Achieves state-of-the-art rendering quality in lane switching scenarios.

Effectively handles large viewpoint changes without multi-lane datasets.

Demonstrates superior performance over existing methods.

Abstract

We propose GGS, a Generalizable Gaussian Splatting method for Autonomous Driving which can achieve realistic rendering under large viewpoint changes. Previous generalizable 3D gaussian splatting methods are limited to rendering novel views that are very close to the original pair of images, which cannot handle large differences in viewpoint. Especially in autonomous driving scenarios, images are typically collected from a single lane. The limited training perspective makes rendering images of a different lane very challenging. To further improve the rendering capability of GGS under large viewpoint changes, we introduces a novel virtual lane generation module into GSS method to enables high-quality lane switching even without a multi-lane dataset. Besides, we design a diffusion loss to supervise the generation of virtual lane image to further address the problem of lack of data in the virtual lanes. Finally, we also propose a depth refinement module to optimize depth estimation in the GSS model. Extensive validation of our method, compared to existing approaches, demonstrates state-of-the-art performance.