MVSNeRF: Fast Generalizable Radiance Field Reconstruction from Multi-View Stereo

TL;DR

MVSNeRF introduces a fast, generalizable neural radiance field reconstruction method from just three views, leveraging multi-view stereo techniques for efficient view synthesis across diverse scenes.

Contribution

It proposes a neural network that reconstructs radiance fields from minimal views using plane-swept cost volumes, enabling rapid, scene-generalizable rendering.

Findings

Outperforms existing methods in view synthesis quality.

Generalizes well across different scene types, including indoor environments.

Enables quick per-scene refinement with less optimization time.

Abstract

We present MVSNeRF, a novel neural rendering approach that can efficiently reconstruct neural radiance fields for view synthesis. Unlike prior works on neural radiance fields that consider per-scene optimization on densely captured images, we propose a generic deep neural network that can reconstruct radiance fields from only three nearby input views via fast network inference. Our approach leverages plane-swept cost volumes (widely used in multi-view stereo) for geometry-aware scene reasoning, and combines this with physically based volume rendering for neural radiance field reconstruction. We train our network on real objects in the DTU dataset, and test it on three different datasets to evaluate its effectiveness and generalizability. Our approach can generalize across scenes (even indoor scenes, completely different from our training scenes of objects) and generate realistic view synthesis results using only three input images, significantly outperforming concurrent works on generalizable radiance field reconstruction. Moreover, if dense images are captured, our estimated radiance field representation can be easily fine-tuned; this leads to fast per-scene reconstruction with higher rendering quality and substantially less optimization time than NeRF.