DebSDF: Delving into the Details and Bias of Neural Indoor Scene Reconstruction

TL;DR

DebSDF addresses the challenge of reconstructing thin, detailed indoor surfaces from multi-view images by integrating uncertainty reasoning for monocular priors with a bias-aware SDF-to-density transformation. It filters unreliable priors using a view-dependent uncertainty map, concentrates ray sampling and adaptive smoothing on high-detail regions, and reduces density rendering bias caused by surface curvature. The method achieves state-of-the-art quantitative and qualitative results across multiple indoor datasets, demonstrating improved reconstruction of thin structures such as chair legs and lampshades. This approach enhances robustness to monocular-prior domain gaps and offers a principled path to more faithful neural implicit reconstructions in texture-less environments.

Abstract

In recent years, the neural implicit surface has emerged as a powerful representation for multi-view surface reconstruction due to its simplicity and state-of-the-art performance. However, reconstructing smooth and detailed surfaces in indoor scenes from multi-view images presents unique challenges. Indoor scenes typically contain large texture-less regions, making the photometric loss unreliable for optimizing the implicit surface. Previous work utilizes monocular geometry priors to improve the reconstruction in indoor scenes. However, monocular priors often contain substantial errors in thin structure regions due to domain gaps and the inherent inconsistencies when derived independently from different views. This paper presents \textbf{DebSDF} to address these challenges, focusing on the utilization of uncertainty in monocular priors and the bias in SDF-based volume rendering. We propose an uncertainty modeling technique that associates larger uncertainties with larger errors in the monocular priors. High-uncertainty priors are then excluded from optimization to prevent bias. This uncertainty measure also informs an importance-guided ray sampling and adaptive smoothness regularization, enhancing the learning of fine structures. We further introduce a bias-aware signed distance function to density transformation that takes into account the curvature and the angle between the view direction and the SDF normals to reconstruct fine details better. Our approach has been validated through extensive experiments on several challenging datasets, demonstrating improved qualitative and quantitative results in reconstructing thin structures in indoor scenes, thereby outperforming previous work.

Figures (12)

• Figure 1: We can observe that our method can reconstruct the indoor scene with more detailed structures, such as the chair legs and bracket of the desk lamp. Previous works such as MonoSDF yu2022monosdf, which is based on the VolSDF yariv2021volume, can not reconstruct the thin and detailed surface due to the inaccurate geometry prior at these regions. Our method can accurately generate the uncertainty map, which can localize the inaccurate priors and reduce the bias in SDF-based rendering with a proposed bias-aware SDF to density transformation approach so that our method can reconstruct the indoor scene significantly better than previous works.
• Figure 2: The rendered uncertainty map can localize the regions where the monocular prior is inaccurate, which usually corresponds to thin structures with high levels of detail. Compared with NeuRIS wang2022neuris, our model is capable of generating more reasonable results.
• Figure 3: The overview of our method. We propose the masked uncertainty learning to adaptively filter the geometry prior and localize the detailed and thin region in 5D space so that the small and thin structure would not be lost due to the wrong prior. Then, the localized uncertainty maps are utilized to guide ray sampling and smooth regularization to improve the reconstruction details of geometry. Besides, we analyze the bias in volume rendering caused by the transformation from SDF to density, which has a significant negative impact on the small and thin structure with geometry prior. A bias-aware SDF to density transformation is proposed to significantly reduce the bias for reconstructing small and thin objects.
• Figure 4: As shown in sub-figure (i), the TUVR zhang2023towards applies the cosine between the ray direction and normal $\cos\theta$ to reduce the bias. Specifically, it transforms the SDF $s$ to the depth $s/|cos\theta|$ ($of\rightarrow od$) by assuming the ray intersects with the planar surface. However, TUVR zhang2023towards, NeuS wang2021neus, and HF-NeuS wang2022hf only consider the planar surface, which ignores the curvature of the surface. As shown in sub-figure (ii), the $dd'$ indicates the error between the rendered depth and the real depth, which causes the biased volume rendering.
• Figure 5: We demonstrate 2 toy cases to simulate the SDF-based rendering by applying Logistic CDF when a ray brushes past and intersects with a small object. The weight function along the ray demonstrates that our method does not have a peak if there is no intersection with the surface. The sub-figures of the right column show the rendered depth corresponding to different $\alpha$ values. Our method can achieve smaller errors than previous works such as the TUVR zhang2023towards and the VolSDF yariv2021volume. We do not demonstrate NeuS wang2021neus since the TUVR zhang2023towards demonstrates that its bias is smaller than NeuS wang2021neus.