面向工业缺陷检测的连续性驱动表征学习
Continuity-Driven Representation Learning for Industrial Defect Detection
- Chung-Ang University（中央大学）
机构由 AI 辅助整理，请以论文原文为准。
AI总结：
针对工业缺陷检测中正常区域表征约束弱的问题，提出连续性驱动表征正则化框架，含多连续性损失与差分损失，在多数据集及检测器上均实现显著性能提升，尤其适配标注数据稀缺场景。
AI中文摘要：
工业缺陷检测与自然图像目标检测存在差异：检测图像在受控条件下采集，包含大量占主导地位的正常区域，且这些区域具有重复结构。因此，缺陷表现为可预测模式的局部中断，而传统检测器主要依赖稀疏的边界框监督，导致正常区域表征的约束较弱。我们提出一种连续性驱动表征正则化框架，将占主导地位的正常区域作为密集辅助监督。该框架引入两个与检测器无关的目标：多连续性损失（Multi-Continuity Loss），结合一维图像块序列预测与二维掩码空间预测；差分损失（Differencing Loss），用于正则化相邻图像块嵌入之间的一阶特征变化和二阶曲率。两个目标均采用基于边界框的区域加权，以稳定正常区域表征，同时保留与缺陷相关的不连续性。在两个真实工业数据集及公开的NEU-DET基准上，使用包括YOLO系列模型、MambaYOLO和DETR在内的六种检测器架构开展实验，结果表明其相较于原生检测器基线均实现了一致提升。在全数据设置下，所提正则化器使平均mAP@0.5:0.95在Industrial Metal数据集上提升最多3.49个百分点，在MEA数据集上提升5.38个百分点，在NEU-DET数据集上提升5.03个百分点。在有限数据条件下，增益更为显著：仅使用25%的训练数据时，差分损失（Differencing Loss）在NEU-DET数据集上实现mAP@0.5提升最多21.07个百分点，mAP@0.5:0.95提升8.23个百分点。这些结果表明，连续性驱动正则化为改进工业缺陷检测提供了有效先验，尤其在标注数据稀缺时效果显著。
英文摘要：
Industrial defect detection differs from natural-image object detection because inspection images are captured under controlled conditions and contain large normal-dominant regions with repetitive structures. Defects therefore appear as localized disruptions of otherwise predictable patterns, while conventional detectors rely mainly on sparse bounding-box supervision, resulting in weakly constrained normal-region representations. We propose a continuity-driven representation regularization framework that exploits normal-dominant regions as dense auxiliary supervision. The framework introduces two detector-agnostic objectives: Multi-Continuity Loss, which combines 1D patch-sequence prediction and 2D masked spatial prediction, and Differencing Loss, which regularizes first-order feature variation and second-order curvature between neighboring patch embeddings. Both objectives are applied with box-derived region weighting to stabilize normal-region representations while preserving defect-related discontinuities. Experiments on two real-world industrial datasets and the public NEU-DET benchmark, using six detector architectures including YOLO-family models, MambaYOLO, and DETR, demonstrate consistent improvements over native detector baselines. In the full-data setting, the proposed regularizers improve average mAP@0.5:0.95 by up to 3.49 percentage points on Industrial Metal, 5.38 percentage points on MEA, and 5.03 percentage points on NEU-DET. Under limited-data conditions, the gains become more pronounced, with Differencing Loss achieving improvements of up to 21.07 percentage points in mAP@0.5 and 8.23 percentage points in mAP@0.5:0.95 on NEU-DET using only 25% of the training data. These results suggest that continuity-driven regularization provides an effective prior for improving industrial defect detection, particularly when annotated data are scarce.
