Search papers, labs, and topics across Lattice.
SAFEVPR is introduced as a novel, non-trainable verification-and-calibration pipeline to ensure safe sequence-based visual place recognition (VPR) under cross-condition deployment. It replaces cosine similarity with a mutual-nearest-neighbor patch-matching score derived from frozen DINOv2 ViT features and employs Mondrian conformal Learn-Then-Test calibration with Bonferroni correction. Empirical validation across 23 cross-condition setups demonstrates SAFEVPR's ability to maintain false-discovery-rate control, achieving a mean accepted FDR of 0.014 and a mean TPR of 0.75 at a target FDR of 0.10.
Even with comparable AUROC, standard VPR methods fail to provide valid conformal prediction guarantees under dataset shift, but SAFEVPR achieves empirical validity across diverse conditions by combining patch-based matching with specialized calibration.
Sequence-based visual place recognition (VPR) for SLAM and robot relocalization must decide whether the retrieved top-1 candidate is safe to accept. Conformal prediction is a natural framework for this accept/reject decision, but its finite-sample guarantees rely on exchangeability between calibration and deployment (test) data, which is violated under cross-condition deployment. We introduce SAFEVPR, a non-trainable verification-and-calibration pipeline for safe cross-condition sequence VPR. SAFEVPR replaces the standard backbone cosine similarity with a mutual-nearest-neighbour (MNN) patch-matching score computed from frozen DINOv2 ViT features, and replaces flat Learn-Then-Test calibration with Mondrian conformal LTT, fitting separate Bonferroni-corrected thresholds across score bins. Under exchangeability, these thresholds would provide finite-sample false-discovery-rate (FDR) control; under condition shift, we evaluate empirical validity per deployment. Across 23 cross-condition setups from Oxford RobotCar, NCLT, and St Lucia datasets, using three frozen VPR backbones, SAFEVPR is empirically valid on 23/23 setups at target FDR alpha = 0.10, achieving mean accepted FDR 0.014 and mean true-positive rate (TPR) 0.75. The results show that raw discrimination alone is not sufficient for conformal validity: AnyLoc-VLAD and Super-Point+LightGlue reach comparable area under the receiver operating characteristic curve (AUROC) but fail more setups under the same calibration. On textureless repetitive scenery, SAFEVPR safely abstains rather than accepting unreliable matches. Code is available at https://github.com/Hasar12139/SafeVPR.
