H2O-SDF: Two-phase Learning for 3D Indoor Reconstruction using Object Surface Fields

TL;DR

H2O-SDF introduces a two-phase learning framework with Object Surface Fields to improve 3D indoor scene reconstruction by balancing room layout preservation and detailed object surfaces.

Contribution

The paper presents a novel two-phase learning method and the Object Surface Field concept to enhance high-frequency detail capture in 3D indoor reconstruction.

Findings

Effective preservation of room geometry.

Enhanced capture of object surface details.

Mitigation of vanishing gradient issues.

Abstract

Advanced techniques using Neural Radiance Fields (NeRF), Signed Distance Fields (SDF), and Occupancy Fields have recently emerged as solutions for 3D indoor scene reconstruction. We introduce a novel two-phase learning approach, H2O-SDF, that discriminates between object and non-object regions within indoor environments. This method achieves a nuanced balance, carefully preserving the geometric integrity of room layouts while also capturing intricate surface details of specific objects. A cornerstone of our two-phase learning framework is the introduction of the Object Surface Field (OSF), a novel concept designed to mitigate the persistent vanishing gradient problem that has previously hindered the capture of high-frequency details in other methods. Our proposed approach is validated through several experiments that include ablation studies.