DGNR: Density-Guided Neural Point Rendering of Large Driving Scenes

TL;DR

DGNR introduces a density-guided neural rendering framework that effectively synthesizes large-scale driving scenes in real-time without requiring geometric priors, outperforming existing methods in quality and efficiency.

Contribution

The paper proposes a novel density-guided neural rendering approach that learns a density space for large scenes, eliminating the need for geometric priors and enabling real-time, high-quality rendering.

Findings

Effective synthesis of photorealistic driving scenes

Real-time rendering performance demonstrated

Outperforms existing large-scale scene rendering methods

Abstract

Despite the recent success of Neural Radiance Field (NeRF), it is still challenging to render large-scale driving scenes with long trajectories, particularly when the rendering quality and efficiency are in high demand. Existing methods for such scenes usually involve with spatial warping, geometric supervision from zero-shot normal or depth estimation, or scene division strategies, where the synthesized views are often blurry or fail to meet the requirement of efficient rendering. To address the above challenges, this paper presents a novel framework that learns a density space from the scenes to guide the construction of a point-based renderer, dubbed as DGNR (Density-Guided Neural Rendering). In DGNR, geometric priors are no longer needed, which can be intrinsically learned from the density space through volumetric rendering. Specifically, we make use of a differentiable renderer to synthesize images from the neural density features obtained from the learned density space. A density-based fusion module and geometric regularization are proposed to optimize the density space. By conducting experiments on a widely used autonomous driving dataset, we have validated the effectiveness of DGNR in synthesizing photorealistic driving scenes and achieving real-time capable rendering.