We present a novel framework for unsupervised object-centric 3D scene understanding that generalizes robustly to out-of-distribution images. To achieve this, we design a causal generative model reflecting the physical process by which an image is produced, when a camera captures a scene containing multiple objects. This model is trained to reconstruct multi-view images via a latent representation describing the shapes, colours and positions of the 3D objects they show. It explicitly represents object instances as separate neural radiance fields, placed into a 3D scene. We then propose an inference algorithm that can infer this latent representation given a single out-of-distribution image as input -- even when it shows an unseen combination of components, unseen spatial compositions or a radically new viewpoint. We conduct extensive experiments applying our approach to test datasets that have zero probability under the training distribution. These show that it accurately reconstructs a scene's geometry, segments objects and infers their positions, despite not receiving any supervision. Our approach significantly out-performs baselines that do not capture the true causal image generation process.

Motivation: Clustering analysis is a key technique for quantitatively characterizing structures in localization microscopy images. To build up accurate information about biological structures, it is critical that the quantification is both accurate (close to the ground truth) and precise (has small scatter and is reproducible). Results: Here, we describe how the Rényi divergence can be used for cluster radius measurements in localization microscopy data. We demonstrate that the Rényi divergence can operate with high levels of background and provides results which are more accurate than Ripley's functions, Voronoi tesselation or DBSCAN. Availability and implementation: The data supporting this research and the software described are accessible at the following site: https://dx.doi.org/10.18742/RDM01-316. Correspondence and requests for materials should be addressed to the corresponding author. Supplementary information: Supplementary data are available at Bioinformatics online.