Distributional Uncertainty for Out-of-Distribution Detection

2025-07-24

AI Summary

Existing out-of-distribution (OoD) detection methods for semantic segmentation often decouple model uncertainty estimation from the semantic misclassification objective, leading to semantically inconsistent uncertainty quantification. Method: We propose the Free Energy Posterior Network (FEPN), which innovatively couples the free energy function with Beta distribution density estimation to construct a differentiable, sampling-free uncertainty parameterization module. Additionally, we introduce a Residual Prediction Branch (RPL) that jointly optimizes OoD region identification and distributional uncertainty estimation. Contribution/Results: FEPN achieves state-of-the-art OoD detection performance on Fishyscapes, RoadAnomaly, and Segment-Me-If-You-Can benchmarks, improving AUPR by 3.2–7.8% while accelerating inference by 2.1×. Crucially, it is the first method to enable end-to-end alignment between uncertainty estimation and semantic error localization, ensuring semantic consistency in uncertainty-aware segmentation.

Technology Category

• Machine Learning: Calibration & Uncertainty Quantification
• Computer Vision: Segmentation
• Reasoning under Uncertainty: Relational Probabilistic Models

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Estimating uncertainty from deep neural networks is a widely used approach for detecting out-of-distribution (OoD) samples, which typically exhibit high predictive uncertainty. However, conventional methods such as Monte Carlo (MC) Dropout often focus solely on either model or data uncertainty, failing to align with the semantic objective of OoD detection. To address this, we propose the Free-Energy Posterior Network, a novel framework that jointly models distributional uncertainty and identifying OoD and misclassified regions using free energy. Our method introduces two key contributions: (1) a free-energy-based density estimator parameterized by a Beta distribution, which enables fine-grained uncertainty estimation near ambiguous or unseen regions; and (2) a loss integrated within a posterior network, allowing direct uncertainty estimation from learned parameters without requiring stochastic sampling. By integrating our approach with the residual prediction branch (RPL) framework, the proposed method goes beyond post-hoc energy thresholding and enables the network to learn OoD regions by leveraging the variance of the Beta distribution, resulting in a semantically meaningful and computationally efficient solution for uncertainty-aware segmentation. We validate the effectiveness of our method on challenging real-world benchmarks, including Fishyscapes, RoadAnomaly, and Segment-Me-If-You-Can.

Problem

• Detecting out-of-distribution samples using uncertainty estimation
• Jointly modeling distributional uncertainty for OoD detection
• Improving semantic segmentation with uncertainty-aware learning

Innovation

• Free-Energy Posterior Network for joint uncertainty modeling
• Beta distribution-based fine-grained uncertainty estimation
• Integrated loss enables direct uncertainty estimation