Region-Aware CAM: High-Resolution Weakly-Supervised Defect Segmentation via Salient Region Perception

2025-06-28

AI Summary

Industrial surface defect detection faces challenges of high annotation costs and low localization accuracy in weakly supervised methods. To address the low spatial resolution and severe detail loss inherent in Class Activation Mapping (CAM) heatmaps, this paper proposes a weakly supervised semantic segmentation framework. It introduces a filter-guided backpropagation mechanism to improve gradient-based localization quality, incorporates a region-aware weighting module to enhance spatial consistency of defect responses, and employs an iterative pseudo-label optimization strategy for high-resolution defect segmentation. Extensive experiments on multiple industrial defect datasets demonstrate that the proposed method significantly outperforms existing weakly supervised approaches and substantially narrows the performance gap with fully supervised segmentation models. The framework is particularly effective in real-world industrial scenarios where labeled data are scarce.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Semi-Supervised Learning
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

Surface defect detection plays a critical role in industrial quality inspection. Recent advances in artificial intelligence have significantly enhanced the automation level of detection processes. However, conventional semantic segmentation and object detection models heavily rely on large-scale annotated datasets, which conflicts with the practical requirements of defect detection tasks. This paper proposes a novel weakly supervised semantic segmentation framework comprising two key components: a region-aware class activation map (CAM) and pseudo-label training. To address the limitations of existing CAM methods, especially low-resolution thermal maps, and insufficient detail preservation, we introduce filtering-guided backpropagation (FGBP), which refines target regions by filtering gradient magnitudes to identify areas with higher relevance to defects. Building upon this, we further develop a region-aware weighted module to enhance spatial precision. Finally, pseudo-label segmentation is implemented to refine the model's performance iteratively. Comprehensive experiments on industrial defect datasets demonstrate the superiority of our method. The proposed framework effectively bridges the gap between weakly supervised learning and high-precision defect segmentation, offering a practical solution for resource-constrained industrial scenarios.

Problem

• Weakly-supervised defect segmentation with limited annotations
• Low-resolution CAM methods lacking detail preservation
• Bridging weakly supervised learning and high-precision segmentation

Innovation

• Region-aware CAM for defect segmentation
• Filtering-guided backpropagation (FGBP) refines target regions
• Pseudo-label training enhances model performance