Direct Voxel Grid Optimization: Super-fast Convergence for Radiance Fields Reconstruction

TL;DR

This paper introduces a rapid convergence method for scene radiance field reconstruction using voxel grids, achieving NeRF-like quality in under 15 minutes with a single GPU.

Contribution

It proposes a novel voxel grid-based approach with post-activation interpolation and priors, enabling fast and high-quality scene reconstruction.

Findings

Achieves NeRF-comparable quality in less than 15 minutes.

Uses explicit voxel representations for geometry and appearance.

Outperforms existing methods in convergence speed and quality.

Abstract

We present a super-fast convergence approach to reconstructing the per-scene radiance field from a set of images that capture the scene with known poses. This task, which is often applied to novel view synthesis, is recently revolutionized by Neural Radiance Field (NeRF) for its state-of-the-art quality and flexibility. However, NeRF and its variants require a lengthy training time ranging from hours to days for a single scene. In contrast, our approach achieves NeRF-comparable quality and converges rapidly from scratch in less than 15 minutes with a single GPU. We adopt a representation consisting of a density voxel grid for scene geometry and a feature voxel grid with a shallow network for complex view-dependent appearance. Modeling with explicit and discretized volume representations is not new, but we propose two simple yet non-trivial techniques that contribute to fast convergence speed and high-quality output. First, we introduce the post-activation interpolation on voxel density, which is capable of producing sharp surfaces in lower grid resolution. Second, direct voxel density optimization is prone to suboptimal geometry solutions, so we robustify the optimization process by imposing several priors. Finally, evaluation on five inward-facing benchmarks shows that our method matches, if not surpasses, NeRF's quality, yet it only takes about 15 minutes to train from scratch for a new scene.