HyperNeRF: A Higher-Dimensional Representation for Topologically Varying Neural Radiance Fields

TL;DR

HyperNeRF introduces a higher-dimensional representation for Neural Radiance Fields, enabling modeling of topological changes in scenes, and outperforms existing methods in scene interpolation and novel-view synthesis.

Contribution

The paper proposes HyperNeRF, a novel higher-dimensional approach that effectively models topological variations in scenes, overcoming limitations of continuous deformation fields.

Findings

HyperNeRF reduces error rates by 4.1% in interpolation.

HyperNeRF reduces error rates by 8.6% in novel-view synthesis.

HyperNeRF outperforms existing methods like Nerfies on both tasks.

Abstract

Neural Radiance Fields (NeRF) are able to reconstruct scenes with unprecedented fidelity, and various recent works have extended NeRF to handle dynamic scenes. A common approach to reconstruct such non-rigid scenes is through the use of a learned deformation field mapping from coordinates in each input image into a canonical template coordinate space. However, these deformation-based approaches struggle to model changes in topology, as topological changes require a discontinuity in the deformation field, but these deformation fields are necessarily continuous. We address this limitation by lifting NeRFs into a higher dimensional space, and by representing the 5D radiance field corresponding to each individual input image as a slice through this "hyper-space". Our method is inspired by level set methods, which model the evolution of surfaces as slices through a higher dimensional surface. We evaluate our method on two tasks: (i) interpolating smoothly between "moments", i.e., configurations of the scene, seen in the input images while maintaining visual plausibility, and (ii) novel-view synthesis at fixed moments. We show that our method, which we dub HyperNeRF, outperforms existing methods on both tasks. Compared to Nerfies, HyperNeRF reduces average error rates by 4.1% for interpolation and 8.6% for novel-view synthesis, as measured by LPIPS. Additional videos, results, and visualizations are available at https://hypernerf.github.io.