中文PipeMFL-240K：面向管道漏磁成像目标检测的大规模数据集与基准
ENPipeMFL-240K: A Large-scale Dataset and Benchmark for Object Detection in Pipeline Magnetic Flux Leakage Imaging
PipeMFL-240K数据集发布，用于磁通量泄漏（MFL）管道检测。该数据集含24万张精细标注图像，填补了大规模公开基准的空白，可公平比较和复现深度学习模型，推动管道缺陷自动检测的可靠性与工业化应用。
arXiv:2602.07044v3 Announce Type: replace Abstract: Pipeline integrity is critical to industrial safety and environmental protection, with Magnetic Flux Leakage (MFL) detection being a primary non-destructive testing technology. Despite the promise of deep learning for automating MFL interpretation, progress toward reliable models has been constrained by the absence of a large-scale public dataset and benchmark, making fair comparison and reproducible evaluation difficult. We introduce \textbf{PipeMFL-240K}, a large-scale, meticulously annotated dataset and benchmark for complex object detecti