Quantile Transfer for Reliable Operating Point Selection in Visual Place Recognition

2026-02-04

AI Summary

This work addresses the challenge in visual place recognition (VPR) where conventional methods rely on manually tuned matching thresholds that struggle to adapt to environmental variations, often resulting in a suboptimal trade-off between precision and recall. To overcome this limitation, the authors propose an adaptive threshold selection mechanism that automatically determines the optimal operating point to maximize recall under a user-specified precision constraint. Leveraging a small set of ground-truth-calibrated trajectories, the method employs quantile normalization of similarity score distributions to derive robust thresholds. A key innovation is the introduction of a quantile transfer strategy, which ensures threshold stability across varying calibration scales and query subsets, enabling seamless generalization to new environments without re-tuning. Extensive experiments across multiple benchmark datasets and state-of-the-art VPR pipelines demonstrate that the proposed approach achieves up to a 25% improvement in recall under high-precision regimes, substantially outperforming existing techniques.

Technology Category

• Machine Learning: Calibration & Uncertainty Quantification
• Computer Vision: Adversarial Attacks & Robustness
• Intelligent Robots: Localization, Mapping, and Navigation

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

Visual Place Recognition (VPR) is a key component for localisation in GNSS-denied environments, but its performance critically depends on selecting an image matching threshold (operating point) that balances precision and recall. Thresholds are typically hand-tuned offline for a specific environment and fixed during deployment, leading to degraded performance under environmental change. We propose a method that, given a user-defined precision requirement, automatically selects the operating point of a VPR system to maximise recall. The method uses a small calibration traversal with known correspondences and transfers thresholds to deployment via quantile normalisation of similarity score distributions. This quantile transfer ensures that thresholds remain stable across calibration sizes and query subsets, making the method robust to sampling variability. Experiments with multiple state-of-the-art VPR techniques and datasets show that the proposed approach consistently outperforms the state-of-the-art, delivering up to 25% higher recall in high-precision operating regimes. The method eliminates manual tuning by adapting to new environments and generalising across operating conditions. Our code will be released upon acceptance.

Problem

• Visual Place Recognition
• Operating Point Selection
• Threshold Adaptation
• Environmental Change
• Precision-Recall Trade-off

Innovation

• Quantile Transfer
• Visual Place Recognition
• Operating Point Selection
• Threshold Calibration
• Similarity Score Normalization