Uncertainty-Aware 3D Position Reconstruction in CZT Drift-Strip Detectors
Authors: Alejandro Valverde Mahou, Michał Kossakowski, Selina R. H. Owe, Rikke S. Klausen, Irfan Kuvvetli, Christian A. Naesseth, Søren Hauberg
Organizations: DTU Compute, Technical University of Denmark, Kgs. Lyngby, Denmark. · DTU Space, Technical University of Denmark, Kgs. Lyngby, Denmark. · Informatics Institute, University of Amsterdam, Amsterdam, Netherlands.
Accurate three-dimensional interaction-position reconstruction is important for CZT drift-strip detectors, but conventional methods can suffer from degraded performance in challenging detector regions, while learned models provide limited information about prediction reliability. We investigate two uncertainty-aware approaches: deep ensembles of point-estimation networks and simulation-based inference (SBI), which estimates a posterior distribution over interaction position. Both approaches are trained on physics-based simulations and evaluated on both synthetic and experimental data. On synthetic data, uncertainty from both methods correlates with positioning error, with SBI showing a substantially stronger relationship. On experimental measurements, the deep ensemble achieves improved z-position resolution, while SBI provides more informative uncertainty estimates that highlight regions of reduced positional information and discrepancies between simulation and the physical detector. These results demonstrate the complementary strengths of ensemble-based reconstruction and posterior inference for uncertainty-aware detector position reconstruction.
Figures & tables
Figure 1: Relationship between predicted uncertainty and positioning error on the synthetic test set for the deep ensemble and SBI. Each point represents an event; the dashed line indicates the 1:1 relationship, while r denotes the Pearson correlation coefficient.
Figure 2: Uncertainty estimates for synthetic-data events under full illumination across the anode ( x – y ) and cathode ( z – y ) dimensions. Left: Uncertainty across anodes in the x – y plane; Right: Uncertainty across cathodes in the z – y plane. Top: ensembles; Bottom: SBI . The corresponding full-range uncertainty estimates are shown in Figure 6 and 7 .
Figure 3: Reconstruction resolution for the (a) x - and (b) z -dimensions using conventional pulse-shape analysis, an ensemble of point-estimate models, and SBI . Left: pulse-shape analysis; Center: ensemble; Right: SBI . The 1D histograms show the distribution of reconstructed positions, while the 2D histograms show the same reconstructions against the depth of illumination ( y -position).
Figure 4: Uncertainty estimates for real-data events under full illumination across the anode ( x – y ) and cathode ( z – y ) dimensions. Left: Uncertainty across anodes in the x – y plane; Right: Uncertainty across cathodes in the z – y plane. Top: Ensemble; Bottom: SBI . Figure 8 and 9
Figure 5: Diagonal slit-beam reconstruction for conventional pulse-shape analysis (first from left), ensemble and its uncertainty (second and third), and SBI and its uncertainty (fourth and fifth), viewed in the x – z plane. The beam crosses both anode and cathode strips, including faulty cathode C12 and dead cathode C4 .
Appendix figures & tables4 assets
Supplementary material from the paper’s appendix.
Appendix
Figure 6: Uncertainty estimates for synthetic-data events under full illumination across the anode dimension. Uncertainty is shown in the x – y plane, with the ensemble at the top and SBI at the bottom.
Figure 7: Uncertainty estimates for synthetic-data events under full illumination across the cathode dimension. Uncertainty is shown in the z – y plane, with the ensemble at the top and SBI at the bottom.
Figure 8: Uncertainty estimates for real-data events under full illumination across the anode dimension. Uncertainty is shown in the x – y plane, with the ensemble at the top and SBI at the bottom.
Figure 9: Uncertainty estimates for real-data events under full illumination across the cathode dimension. Uncertainty is shown in the z – y plane, with the ensemble at the top and SBI at the bottom.
Geometric foundation models hold promise for unconstrained dense geometry prediction from uncalibrated images. However, in current feed-forward designs, their predicted confidence scores are heuristic, lack probabilistic interpretation, and often fail to indicate where and how much the predicted geometry can be trusted. To address this gap, we present Trust3R, a lightweight evidential uncertainty framework for feed-forward 3D reconstruction. Trust3R combines gated residual mean refinement with a Normal-Inverse-Wishart evidential head, yielding a closed-form multivariate Student-t distribution for per-point geometric uncertainty. This design provides probabilistically grounded pointmap uncertainty estimates while adding moderate inference overhead. We evaluate on diverse indoor and outdoor benchmarks and compare against MASt3R's built-in confidence map as well as common uncertainty-aware baselines spanning single-pass heteroscedastic regression and sampling-based methods such as MC dropout and deep ensembles. Experimental results show that Trust3R consistently improves risk-coverage and sparsification, and generally improves geometric accuracy. These gains are reflected in stronger uncertainty ranking across benchmarks, with 25% lower AURC and 41% lower AUSE on ScanNet++, providing a practical reliability signal for uncertainty-aware weighting in downstream geometry pipelines. The project page and code are available at https://trust3r-z.github.io/.
Zihao Zhu, Wenyuan Zhao, Nuo Chen +2
Department of Electrical and Computer Engineering, Texas A&M University, College Station, TX, USA
Feed-forward 3D reconstruction models output a per-pixel confidence that is used by downstream systems as an uncertainty signal. The confidence is trained to serve as a weight in the training loss of models. Whether the confidence can be used as an uncertainty magnitude has not been measured. We audit seven backbones on 13 datasets and score the confidence on four properties, i.e., ranking of error, ratio of error to uncertainty on average, slope of this ratio across the confidence range, and coverage of the implied error distribution. Although the confidence ranks error quite well, the uncertainty decoded from the confidence is too small compared to the actual error. The uncertainty has the right size only under the exact training conditions. The median case is off by at least 2.4x across all seven models, while the uncertainty is further off the more confident the model is. Our work shows that the overconfidence appears on unseen scenes even when the model reaches its loss's optimum. As a post-hoc repair we fit a power law on the confidence with two constants per backbone--dataset pair. The repair brings all four audited properties to target at the dataset level, while leaving ranking untouched. Fitted with the target dataset held out, the constants bring the median case from 2.4x off to 1.35x. The repair does not hold below the dataset level, where two-thirds of held-out scenes are still more than five points off in coverage. We attribute what the repair cannot reach to the model, which carries neither the scale of the error nor the shape of its distribution across predictions. We release the audit protocol, its results, and the fitted constants per backbone-dataset pair.
Nanxing Nick Deng, Qing Cheng, Niclas Zeller +1
Technical University of Munich · Munich Center for Machine Learning (MCML) · Karlsruhe University of Applied Sciences
A correctly reconstructed distant point appears uncertain even when a frozen 3D model processes equivalent inputs because its output frame rotates fractionally. This exposes a weakness of trainingfree perturbation uncertainty: when outputs contain an unobserved symmetry, run-to-run variation potentially reflects symmetry rather than error. Existing alternatives have trade-offs: built-in confidence is outperformed in most evaluated conditions, while trained evidential heads require modelspecific supervision. For point maps, this research derives a closed-form, error-independent variance term that grows with scene extent and potentially overwhelms the desired signal. Simulation reproduces the effect; all 30 real VGGT view-sets tested exhibit its predicted ∥xp∥2 signature. ReVar3R robustly registers predictions to a common similarity frame before computing per-point variance, without retraining or modifying the frozen model. Optional calibration and fusion use a held-out split. Across VGGT, π3, and MASt3R on six datasets, the same estimator on every backbone lowers AUSE below built-in confidence in 15 of 18 conditions. The staged evaluation yields 11 of 18 wins for the label-free core, 12/18 for label-free equal-weight fusion, 14/18 with held-out weights, and 15/18 when the built-in signal is included. Against a trained evidential head, the result is a trade-off: the head calibrates magnitude better and leads in its training domain, whereas ReVar3R transfers across backbones without adaptation. Its ranking improves point filtering, but it does not detect stable systematic bias, aid novel-view synthesis, or transfer calibration across domains.
Sammam Mahdi, Fariha Binta Salim, Rakin Bin Rabbani +1