Adaptive Scoring and Thresholding with Human Feedback for Robust Out-of-Distribution Detection

TL;DR

This paper introduces a human-in-the-loop framework that adaptively updates scoring functions and thresholds for out-of-distribution detection, ensuring high true positive rates while controlling false positive rates in real-time.

Contribution

It presents a novel adaptive method with theoretical FPR guarantees that improves OOD detection performance over static approaches.

Findings

Outperforms existing methods on OpenOOD benchmarks.

Achieves higher TPRs with FPR control in dynamic OOD scenarios.

Provides theoretical FPR control guarantees under stationary conditions.

Abstract

Machine Learning (ML) models are trained on in-distribution (ID) data but often encounter out-of-distribution (OOD) inputs during deployment -- posing serious risks in safety-critical domains. Recent works have focused on designing scoring functions to quantify OOD uncertainty, with score thresholds typically set based solely on ID data to achieve a target true positive rate (TPR), since OOD data is limited before deployment. However, these TPR-based thresholds leave false positive rates (FPR) uncontrolled, often resulting in high FPRs where OOD points are misclassified as ID. Moreover, fixed scoring functions and thresholds lack the adaptivity needed to handle newly observed, evolving OOD inputs, leading to sub-optimal performance. To address these challenges, we propose a human-in-the-loop framework that \emph{safely updates both scoring functions and thresholds on the fly} based on real-world OOD inputs. Our method maximizes TPR while strictly controlling FPR at all times, even as the system adapts over time. We provide theoretical guarantees for FPR control under stationary conditions and present extensive empirical evaluations on OpenOOD benchmarks to demonstrate that our approach outperforms existing methods by achieving higher TPRs while maintaining FPR control.