<b>使用 Mamba 增强型 HRNet 增强桥梁损伤检测以进行语义分割</b>
收藏资源简介:
在多样性、复杂性和多视角方面创建一个具有挑战性的数据集。该数据集集成了日本土木工程师学会桥梁损伤数据集[1]、波鸿裂缝数据集[2]、CRACK500数据集[3, 4]。为了进行系统的训练和评估,整个数据集分别以 8:1:1 的比例分为训练集、验证集和测试集。[1] Tonan FUJISHIMA, et al. 使用背景图像额外学习桥接多重损伤分割和 3D 损伤模型生成。人工智能与数据科学 2023;4(3):705–714。[2] Çelik F, König M.一个 sigmoid 优化的编码器-解码器网络,用于通过复制-编辑-粘贴迁移学习进行裂纹分割。计算机辅助土木与基础设施工程 2022;37(14):1875–1890 年。[3] Zhang L, Yang F, et al. 使用深度卷积神经网络进行道路裂缝检测.在:2016 年 IEEE 图像处理国际会议 (ICIP) IEEE;2016. 第 3708-3712 页–。[4] Yang F, Zhang L, et al. 用于路面裂缝检测的特征金字塔和分层提升网络.IEEE Transactions on Intelligent Transportation Systems 2019;21(4):1525–1535。
This challenging dataset is constructed with considerations of diversity, complexity and multi-perspective attributes. It integrates the Bridge Damage Dataset from the Japan Society of Civil Engineers (JSCE) [1], the Bochum Crack Dataset [2], and the CRACK500 Dataset [3, 4]. For systematic training and evaluation, the entire dataset is split into training, validation and test sets at a ratio of 8:1:1, respectively. [1] Tonan FUJISHIMA, et al. Additional Learning of Bridge Multi-Damage Segmentation and 3D Damage Model Generation Using Background Images. Artificial Intelligence and Data Science 2023; 4(3): 705–714. [2] Çelik F, König M. A Sigmoid-Optimized Encoder-Decoder Network for Crack Segmentation via Transfer Learning with Copy-Edit-Paste. Computer-Aided Civil and Infrastructure Engineering 2022; 37(14): 1875–1890. [3] Zhang L, Yang F, et al. Road Crack Detection Using Deep Convolutional Neural Networks. In: 2016 IEEE International Conference on Image Processing (ICIP); IEEE; 2016. pp. 3708–3712. [4] Yang F, Zhang L, et al. Feature Pyramid and Hierarchical Boosting Network for Pavement Crack Detection. IEEE Transactions on Intelligent Transportation Systems 2019; 21(4): 1525–1535.




