UnGAP: Uncertainty-Guided Affine Prompting for Real-Time Crack Segmentation

2026-05-04

AI Summary

This work addresses the limitations of existing real-time crack segmentation methods, which struggle to handle aleatoric uncertainty caused by factors such as illumination variations and motion blur, and fail to leverage uncertainty information to refine feature representations. To overcome this, the authors propose an uncertainty-driven closed-loop learning framework that estimates pixel-wise heteroscedastic aleatoric uncertainty and introduces an Uncertainty-Prompted Feature Modulator (UPFM). The UPFM dynamically recalibrates feature distributions by transforming gradients from high-uncertainty regions—typically suppressed in conventional approaches—into affine prompting signals. Additionally, a boundary-aware detection head is incorporated to enhance segmentation accuracy along crack boundaries. The proposed method achieves significant improvements in segmentation performance while maintaining real-time inference speed, demonstrating the efficacy of repurposing uncertainty from a passive diagnostic metric into an active calibration mechanism.

Abstract

Real-time crack segmentation is vital for structural health monitoring but is plagued by aleatoric uncertainties arising from varying lighting, blur, and texture ambiguity. Current uncertainty-aware approaches typically treat uncertainty estimation as a passive endpoint for post-hoc analysis, failing to close the loop by feeding this information back to refine feature representations. We contend that independent pixel-wise heteroscedastic modeling is uniquely suited for crack segmentation, as cracks are defined by fine-grained local gradients rather than the global semantic coherence relied upon in general object segmentation. However, this approach suffers from a structural optimization pathology: high predicted variance attenuates loss gradients, effectively causing the model to ignore difficult samples and under-fit complex boundaries. To address these challenges, we propose UnGAP, a novel framework that establishes a closed-loop mechanism between uncertainty estimation and feature learning. Central to our approach is the Uncertainty-Prompted Feature Modulator (UPFM), which treats aleatoric uncertainty as an active visual prompt rather than a mere output. UPFM dynamically calibrates feature distributions through pixel-wise affine transformations. Crucially, this mechanism mitigates the heteroscedastic pathology by transforming high variance, which would otherwise indicate gradient suppression, into a constructive signal for stronger feature rectification in ambiguous regions. Additionally, a boundary-aware detection head is introduced to further constrain prediction precision. Extensive experiments demonstrate that UnGAP balances superior segmentation accuracy with real-time inference speed, effectively validating the benefit of transforming uncertainty from a passive metric into an active calibration tool.

Problem

• real-time crack segmentation
• aleatoric uncertainty
• heteroscedastic modeling
• structural health monitoring
• uncertainty estimation

Innovation

• Uncertainty-Guided Prompting
• Aleatoric Uncertainty
• Heteroscedastic Modeling
• Affine Feature Modulation
• Real-Time Crack Segmentation