SparseNeuS: Fast Generalizable Neural Surface Reconstruction from Sparse Views

TL;DR

SparseNeuS is a neural surface reconstruction method that effectively works with sparse multi-view images, generalizes to new scenes, and outperforms existing approaches in quality and efficiency.

Contribution

It introduces a novel framework combining geometry encoding volumes and multi-level reasoning for fast, generalizable surface reconstruction from sparse views.

Findings

Outperforms state-of-the-art methods in sparse-view reconstruction

Works well with as few as 2-3 images per scene

Demonstrates good efficiency, generalizability, and flexibility

Abstract

We introduce SparseNeuS, a novel neural rendering based method for the task of surface reconstruction from multi-view images. This task becomes more difficult when only sparse images are provided as input, a scenario where existing neural reconstruction approaches usually produce incomplete or distorted results. Moreover, their inability of generalizing to unseen new scenes impedes their application in practice. Contrarily, SparseNeuS can generalize to new scenes and work well with sparse images (as few as 2 or 3). SparseNeuS adopts signed distance function (SDF) as the surface representation, and learns generalizable priors from image features by introducing geometry encoding volumes for generic surface prediction. Moreover, several strategies are introduced to effectively leverage sparse views for high-quality reconstruction, including 1) a multi-level geometry reasoning framework to recover the surfaces in a coarse-to-fine manner; 2) a multi-scale color blending scheme for more reliable color prediction; 3) a consistency-aware fine-tuning scheme to control the inconsistent regions caused by occlusion and noise. Extensive experiments demonstrate that our approach not only outperforms the state-of-the-art methods, but also exhibits good efficiency, generalizability, and flexibility.