Self-Supervised Visual Prompting for Cross-Domain Road Damage Detection

2025-11-15

AI Summary

Weak cross-domain generalization severely hinders the practical deployment of automated pavement defect detection. To address this, we propose the first unsupervised self-supervised visual prompting framework tailored for road damage detection: it generates defect-aware visual prompts on unlabeled target-domain images to guide representation adaptation of a frozen Vision Transformer (ViT) backbone. We introduce two key innovations—the Self-supervised Prompt Enhancement Module (SPEM) and the Domain-Aware Prompt Alignment (DAPA) strategy—enabling focused defect feature learning and cross-domain representation alignment. Our method requires no annotations, learning prompts solely from target-domain imagery. Evaluated on four benchmark datasets, it achieves superior zero-shot transfer performance over state-of-the-art supervised, self-supervised, and domain-adaptation methods, while significantly improving few-shot adaptation efficiency and cross-domain robustness.

Technology Category

• Computer Vision: Adversarial Attacks & Robustness
• Machine Learning: Unsupervised & Self-Supervised Learning
• Natural Language Processing: Prompt Engineering / Prompting

Application Category

• Search and Retrieval-Augmented AI: Vertical and domain-specific search
• Responsible Web: Data and user privacy-enhancing technologies for the Web
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs

Abstract

The deployment of automated pavement defect detection is often hindered by poor cross-domain generalization. Supervised detectors achieve strong in-domain accuracy but require costly re-annotation for new environments, while standard self-supervised methods capture generic features and remain vulnerable to domain shift. We propose ours, a self-supervised framework that emph{visually probes} target domains without labels. ours introduces a Self-supervised Prompt Enhancement Module (SPEM), which derives defect-aware prompts from unlabeled target data to guide a frozen ViT backbone, and a Domain-Aware Prompt Alignment (DAPA) objective, which aligns prompt-conditioned source and target representations. Experiments on four challenging benchmarks show that ours consistently outperforms strong supervised, self-supervised, and adaptation baselines, achieving robust zero-shot transfer, improved resilience to domain variations, and high data efficiency in few-shot adaptation. These results highlight self-supervised prompting as a practical direction for building scalable and adaptive visual inspection systems. Source code is publicly available: https://github.com/xixiaouab/PROBE/tree/main

Problem

• Addressing poor cross-domain generalization in automated pavement defect detection
• Eliminating costly re-annotation requirements for new environmental domains
• Improving resilience to domain shift in visual inspection systems

Innovation

• Self-supervised prompting for cross-domain defect detection
• Defect-aware prompts derived from unlabeled target data
• Domain-aligned prompt conditioning for robust transfer