Abstract
Reliable uncertainty estimation is essential for deploying object detectors when distribution/covariate shift and adversarial attacks may occur. Existing approaches often require detector retraining, architectural modification, or repeated inference, which may be infeasible or incur significant overheads. We introduce a lightweight post-hoc evidential meta-model that learns when object localisations should be considered uncertain while keeping the base detector frozen. Our approach automatically identifies localisation-relevant features and uses saliency-guided modification to construct an increasingly challenging curriculum. Detection-level targets combine localisation error, modification level, and prediction instability to guide an evidential meta-model to estimate uncertainty for each predicted bounding box. Our approach requires no changes to the detector and preserves its original localisation outputs. Across adversarial attacks and evaluated strengths, GRACE improves TP-FP AUROC by 22% relative to the strongest comparator in some cases while maintaining in-distribution detection performance.
Keywords
Subject
Publication details
- Journal
- Not available
- Open access
- Green open access
Cite this article
APA 7
Barker, C., Bethell, D., & Gerasimou, S. (2026). Localisation-Aware Uncertainty for Pretrained Object Detection. https://omanscience.com/en/articles/localisation-aware-uncertainty-for-pretrained-object-detection
MLA 9
Barker, Charmaine, et al. "Localisation-Aware Uncertainty for Pretrained Object Detection." https://omanscience.com/en/articles/localisation-aware-uncertainty-for-pretrained-object-detection.
Chicago (author–date)
Barker, Charmaine, Daniel Bethell, and Simos Gerasimou. 2026. "Localisation-Aware Uncertainty for Pretrained Object Detection." https://omanscience.com/en/articles/localisation-aware-uncertainty-for-pretrained-object-detection.
Harvard
Barker, C., Bethell, D. and Gerasimou, S. (2026) 'Localisation-Aware Uncertainty for Pretrained Object Detection', Available at: https://omanscience.com/en/articles/localisation-aware-uncertainty-for-pretrained-object-detection.
Vancouver
Barker C, Bethell D, Gerasimou S. Localisation-Aware Uncertainty for Pretrained Object Detection. https://omanscience.com/en/articles/localisation-aware-uncertainty-for-pretrained-object-detection
IEEE
C. Barker, D. Bethell, and S. Gerasimou, "Localisation-Aware Uncertainty for Pretrained Object Detection," https://omanscience.com/en/articles/localisation-aware-uncertainty-for-pretrained-object-detection.