SupLID: Geometrical Guidance for Out-of-Distribution Detection in Semantic Segmentation

2025-11-24

AI Summary

Pixel-wise out-of-distribution (OOD) detection in semantic segmentation suffers from overconfidence of conventional confidence-based methods (e.g., energy, entropy) and insufficient exploitation of geometric structure. To address this, we propose SupLID—a novel post-hoc calibration framework that introduces Linear Intrinsic Dimensionality (LID) to pixel-level OOD detection for the first time. SupLID models the geometric structure of the semantic feature space to construct a geometric core set and designs a superpixel-level fusion scoring mechanism that jointly leverages LID estimates, geometric priors, and classifier confidence. Importantly, it requires no modification to the backbone architecture and is compatible with arbitrary semantic segmentation networks, preserving spatial coherence and enabling efficient inference. Evaluated on standard benchmarks, SupLID achieves state-of-the-art performance across key metrics—including AUR, FPR95, and AUPR—while supporting real-time deployment. The code is publicly available.

Technology Category

• Machine Learning: Calibration & Uncertainty Quantification
• Computer Vision: Segmentation
• Search and Optimization: Learning to Search

Application Category

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

Abstract

Out-of-Distribution (OOD) detection in semantic segmentation aims to localize anomalous regions at the pixel level, advancing beyond traditional image-level OOD techniques to better suit real-world applications such as autonomous driving. Recent literature has successfully explored the adaptation of commonly used image-level OOD methods--primarily based on classifier-derived confidence scores (e.g., energy or entropy)--for this pixel-precise task. However, these methods inherit a set of limitations, including vulnerability to overconfidence. In this work, we introduce SupLID, a novel framework that effectively guides classifier-derived OOD scores by exploiting the geometrical structure of the underlying semantic space, particularly using Linear Intrinsic Dimensionality (LID). While LID effectively characterizes the local structure of high-dimensional data by analyzing distance distributions, its direct application at the pixel level remains challenging. To overcome this, SupLID constructs a geometrical coreset that captures the intrinsic structure of the in-distribution (ID) subspace. It then computes OOD scores at the superpixel level, enabling both efficient real-time inference and improved spatial smoothness. We demonstrate that geometrical cues derived from SupLID serve as a complementary signal to traditional classifier confidence, enhancing the model's ability to detect diverse OOD scenarios. Designed as a post-hoc scoring method, SupLID can be seamlessly integrated with any semantic segmentation classifier at deployment time. Our results demonstrate that SupLID significantly enhances existing classifier-based OOD scores, achieving state-of-the-art performance across key evaluation metrics, including AUR, FPR, and AUP. Code is available at https://github.com/hdnugit/SupLID.

Problem

• Detects anomalous regions at pixel level for semantic segmentation
• Overcomes limitations of classifier-based OOD detection methods
• Integrates geometrical guidance with traditional confidence scores

Innovation

• Uses Linear Intrinsic Dimensionality for geometric guidance
• Constructs geometrical coreset for in-distribution subspace structure
• Computes OOD scores at superpixel level for efficiency