VM-NeRF: Tackling Sparsity in NeRF with View Morphing

TL;DR

VM-NeRF enhances neural radiance field models by generating geometrically consistent intermediate views through view morphing, significantly improving novel view synthesis in sparse data scenarios without prior scene knowledge.

Contribution

The paper introduces VM-NeRF, a novel view morphing-based method that improves NeRF performance with limited viewpoints without requiring scene prior knowledge.

Findings

Up to 1.8dB PSNR improvement with 8 views

Up to 1.0dB PSNR improvement with 4 views

Consistent performance gains over existing sparse-view NeRF methods

Abstract

NeRF aims to learn a continuous neural scene representation by using a finite set of input images taken from various viewpoints. A well-known limitation of NeRF methods is their reliance on data: the fewer the viewpoints, the higher the likelihood of overfitting. This paper addresses this issue by introducing a novel method to generate geometrically consistent image transitions between viewpoints using View Morphing. Our VM-NeRF approach requires no prior knowledge about the scene structure, as View Morphing is based on the fundamental principles of projective geometry. VM-NeRF tightly integrates this geometric view generation process during the training procedure of standard NeRF approaches. Notably, our method significantly improves novel view synthesis, particularly when only a few views are available. Experimental evaluation reveals consistent improvement over current methods that handle sparse viewpoints in NeRF models. We report an increase in PSNR of up to 1.8dB and 1.0dB when training uses eight and four views, respectively. Source code: \url{https://github.com/mbortolon97/VM-NeRF}