Detecting OOD Samples via Optimal Transport Scoring Function

2025-02-22

AI Summary

This work addresses the unsupervised out-of-distribution (OOD) detection problem for deep models during inference under open-world deployment. We propose a model-agnostic, training-free post-hoc OOD scoring method. For the first time, optimal transport theory is introduced into OOD detection: our approach jointly models the geometric structures of feature space, logits space, and softmax probability space to construct a unified multi-space scoring function, backed by theoretical guarantees. Evaluated on CIFAR-10 and CIFAR-100 benchmarks, the method achieves new state-of-the-art performance—reducing false positive rate at 95% true positive rate (FPR@95) by 7.19% and 12.51% respectively when using ResNet-18 and WideResNet-28 backbones. These results demonstrate both effectiveness and strong generalization across architectures.

Technology Category

• Machine Learning: Calibration & Uncertainty Quantification
• Computer Vision: Other Foundations of Computer Vision
• Search and Optimization: Learning to Search

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Graph Algorithms and Modeling for the Web: Algorithms and analysis for incomplete, noisy, or partially observed Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

To deploy machine learning models in the real world, researchers have proposed many OOD detection algorithms to help models identify unknown samples during the inference phase and prevent them from making untrustworthy predictions. Unlike methods that rely on extra data for outlier exposure training, post hoc methods detect Out-of-Distribution (OOD) samples by developing scoring functions, which are model agnostic and do not require additional training. However, previous post hoc methods may fail to capture the geometric cues embedded in network representations. Thus, in this study, we propose a novel score function based on the optimal transport theory, named OTOD, for OOD detection. We utilize information from features, logits, and the softmax probability space to calculate the OOD score for each test sample. Our experiments show that combining this information can boost the performance of OTOD with a certain margin. Experiments on the CIFAR-10 and CIFAR-100 benchmarks demonstrate the superior performance of our method. Notably, OTOD outperforms the state-of-the-art method GEN by 7.19% in the mean FPR@95 on the CIFAR-10 benchmark using ResNet-18 as the backbone, and by 12.51% in the mean FPR@95 using WideResNet-28 as the backbone. In addition, we provide theoretical guarantees for OTOD. The code is available in https://github.com/HengGao12/OTOD.

Problem

• Develops OTOD for OOD detection
• Utilizes optimal transport theory
• Improves FPR@95 performance significantly

Innovation

• Optimal transport theory scoring
• Feature, logit, softmax integration
• Superior CIFAR benchmark performance