ScanNeRF: a Scalable Benchmark for Neural Radiance Fields

In this paper, we propose the first-ever real benchmark thought for evaluating Neural Radiance Fields (NeRFs) and, in general, Neural Rendering (NR) frameworks. We design and implement an effective pipeline for scanning real objects in quantity and effortlessly. Our scan station is built with less than 500$ hardware budget and can collect roughly 4000 images of a scanned object in just 5 minutes. Such a platform is used to build ScanNeRF, a dataset characterized by several train/val/test splits aimed at benchmarking the performance of modern NeRF methods under different conditions. Accordingly, we evaluate three cuttingedge NeRF variants on it to highlight their strengths and weaknesses. The dataset is available on our project page, together with an online benchmark to foster the development of better and better NeRFs.

1. Introduction

What is the Metaverseƒ Stephenson coined this portmanteau in his novel Snow Crash, hypothesizing that in the 21st century humans, thanks to goggles, would be immersed in virtual worlds mixed with real ones. And here we are! At the time, however, the technology to realize the Metaverse was still hypothetical, but today Cross Reality (XR or Extended Reality) is a fact. XR is made up of a multitude of technologies and variants, such as Virtual Reality and Augmented Reality, but they all share a single paradigm: seamless interaction between virtual environments, digital objects and people. That is the Metaverse! But it does not exist yet, and all that is Digital is often only a virtual representation of the real world. How much will it cost us, then, to transport all our real world into the virtual oneƒ