SuperNormal: Neural Surface Reconstruction via Multi-View Normal Integration

TL;DR

SuperNormal is a fast, high-fidelity neural surface reconstruction method that uses multi-view normal integration and innovative gradient approximation techniques to produce detailed 3D surfaces efficiently.

Contribution

It introduces a novel approach combining multi-view normal integration with efficient gradient approximation for neural SDFs, significantly improving speed and detail in 3D reconstruction.

Findings

SuperNormal achieves comparable quality to 3D scanners in minutes.

It is nearly twice as efficient as methods using analytical gradients.

SuperNormal outperforms existing neural reconstruction methods in detail and accuracy.

Abstract

We present SuperNormal, a fast, high-fidelity approach to multi-view 3D reconstruction using surface normal maps. With a few minutes, SuperNormal produces detailed surfaces on par with 3D scanners. We harness volume rendering to optimize a neural signed distance function (SDF) powered by multi-resolution hash encoding. To accelerate training, we propose directional finite difference and patch-based ray marching to approximate the SDF gradients numerically. While not compromising reconstruction quality, this strategy is nearly twice as efficient as analytical gradients and about three times faster than axis-aligned finite difference. Experiments on the benchmark dataset demonstrate the superiority of SuperNormal in efficiency and accuracy compared to existing multi-view photometric stereo methods. On our captured objects, SuperNormal produces more fine-grained geometry than recent neural 3D reconstruction methods.