DeltaSeg通过分层注意力与深度差值学习，精准分割多种结构缺陷。
DeltaSeg: Tiered Attention and Deep Delta Learning for Multi-Class Structural Defect Segmentation
- 采用分层注意力机制，融合通道、坐标与深度差值注意力。
- 在7类和9类缺陷数据集上均超越12种主流模型，平均性能领先。
- 适合需要高精度边界分割的工业缺陷检测场景。
从视觉检测图像中自动分割结构缺陷仍具挑战，源于损伤类型多样、类别极度不平衡及精确边界划分需求。本文提出DeltaSeg，一种带分层注意力策略的U型编码器-解码器架构，集成编码器中的Squeeze-and-Excitation（SE）通道注意力、瓶颈与解码器中的坐标注意力，以及跳跃连接中新颖的深度差值注意力（DDA）模块。编码器使用带空洞阶段的深度可分离卷积，在保持空间分辨率的同时扩大感受野。瓶颈处的空洞空间金字塔池化（ASPP）捕获多尺度上下文。DDA模块通过双路径设计，结合学习到的差值算子抑制干扰特征，并以解码器信号为条件的空间注意力门进行优化。多尺度辅助头实现深层监督，强化梯度流动并促进解码器中间阶段语义特征生成。我们在两个数据集上评估：S2DS（7类）和涵管-污水管道缺陷数据集（CSDD，9类）。在两个基准上，DeltaSeg始终优于12种竞争架构，包括U-Net、SA-UNet、UNet3+、SegFormer、Swin-UNet、EGE-UNet、FPN和Mobile-UNETR，展现出对损伤类型、成像条件和结构几何的强泛化能力。
原文摘要 · Abstract (English)
Automated segmentation of structural defects from visual inspection imagery remains challenging due to the diversity of damage types, extreme class imbalance, and the need for precise boundary delineation. This paper presents DeltaSeg, a U-shaped encoder-decoder architecture with a tiered attention strategy that integrates Squeeze-and-Excitation (SE) channel attention in the encoder, Coordinate Attention at the bottleneck and decoder, and a novel Deep Delta Attention (DDA) mechanism in the skip connections. The encoder uses depthwise separable convolutions with dilated stages to maintain spatial resolution while expanding the receptive field. Atrous Spatial Pyramid Pooling (ASPP) at the bottleneck captures multi-scale context. The DDA module refines skip connections through a dual-path scheme combining a learned delta operator for nuisance feature suppression with spatial attention gates conditioned on decoder signals. Deep supervision through multi-scale auxiliary heads further strengthens gradient flow and encourages semantically meaningful features at intermediate decoder stages. We evaluate DeltaSeg on two datasets: the S2DS dataset (7 classes) and the Culvert-Sewer Defect Dataset (CSDD, 9 classes). Across both benchmarks, DeltaSeg consistently outperforms 12 competing architectures including U-Net, SA-UNet, UNet3+, SegFormer, Swin-UNet, EGE-UNet, FPN, and Mobile-UNETR, demonstrating strong generalization across damage types, imaging conditions, and structural geometries.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。