Density uncertainty quantification with NeRF-Ensembles: Impact of data and scene constraints

In the fields of computer graphics, computer vision and photogrammetry, Neural Radiance Fields (NeRFs) are a major topic driving current research and development. However, the quality of NeRF-generated 3D scene reconstructions and subsequent surface reconstructions, heavily relies on the network output, particularly the density. Regarding this critical aspect, we propose to utilize NeRF-Ensembles that provide a density uncertainty estimate alongside the mean density. We demonstrate that data constraints such as low-quality images and poses lead to a degradation of the rendering quality, increased density uncertainty and decreased predicted density. Even with high-quality input data, the density uncertainty varies based on scene constraints such as acquisition constellations, occlusions and material properties. NeRF-Ensembles not only provide a tool for quantifying the uncertainty but exhibit two promising advantages: Enhanced robustness and artifact removal. Through the mean densities, small outliers are removed, yielding a smoother output with improved completeness. Furthermore, applying a density uncertainty-guided artifact removal in post-processing proves effective for the separation of object and artifact areas. We conduct our methodology on 3 different datasets: (i) synthetic benchmark dataset, (ii) real benchmark dataset, (iii) real data under realistic recording conditions and sensors. • Data constraints (image/pose noise) decrease the NeRF’s density and increase its uncertainty. • Scene constraints (acquisition constellations, occlusions, materials) affect uncertainties. • NeRF-Ensemble mean predictions enhance robustness by removing outliers, smoothing, and improving completeness. • Artifacts (floater) in NeRF predictions are eliminated by according to their density uncertainty.