Don't Just Impute, Detect: Task-Aware Diffusion Pipelines for Edge Anomaly Detection

Computer Science > Machine Learning arXiv:2610.40732 (cs) [Submitted on 29 Oct 2026] Title: Don’t Just Impute, Detect: Task-Aware Diffusion Pipelines for Edge Anomaly Detection Authors: Jeongsu Park, Taehyeon Yun, Geonwoo Kim, Yuji Lim, Moohong Min.

计算机科学 > 机器学习 arXiv:2610.40732 (cs) [提交于 2026 年 10 月 29 日] 标题：不仅是填补，更是检测：用于边缘异常检测的任务感知扩散流水线 作者：Jeongsu Park, Taehyeon Yun, Geonwoo Kim, Yuji Lim, Moohong Min。

Abstract: Industrial Internet of Things (IIoT) edge pipelines routinely lose contiguous blocks of multi-sensor data to network dropouts and hardware faults. The standard remedy is a disjointed “impute-then-detect” pipeline whose imputer is trained only for Mean Squared Error (MSE). Such an imputer fills a missing block with the most probable normal pattern: a partially observed anomaly is smoothed back toward normality, and boundary artifacts on normal data raise false alarms.

摘要：工业物联网 (IIoT) 边缘流水线经常因网络中断和硬件故障而丢失连续的多传感器数据块。标准的补救措施是采用一种脱节的“先填补后检测”流水线，其填补器仅针对均方误差 (MSE) 进行训练。这种填补器会用最可能的正常模式填充缺失块：部分观测到的异常会被平滑回正常状态，而正常数据上的边界伪影则会引发误报。

We call this imputation bias and propose the Task-Aware Diffusion Refiner (TADR), a conditional diffusion refiner trained against the loss landscape of a frozen downstream detector. Because back-propagating through the sampling trajectory is intractable, the task-aware terms act on the one-step clean-window estimate that the noise prediction already yields, so training costs no more than plain diffusion training.

我们将此称为填补偏差，并提出了任务感知扩散精炼器 (TADR)，这是一种针对冻结的下游检测器的损失函数进行训练的条件扩散精炼器。由于通过采样轨迹进行反向传播是不可行的，任务感知项作用于噪声预测已经产生的单步清晰窗口估计上，因此训练成本并不比普通扩散训练高。

Under strict Point-to-Point evaluation over five seeds with paired significance tests, TADR raises F1 by 0.338 over the unrefined detector on SMD (p = 0.0015), and the anomaly enhancement objective alone contributes 0.463 in ablation. We also characterise when the method helps: the gain tracks the continuous, mutually redundant channels available and the anomaly evidence the outage removes, and it is absent on telemetry datasets dominated by one-hot command channels and on PSM.

在五个随机种子下进行严格的点对点评估及配对显著性检验后，TADR 在 SMD 数据集上将 F1 分数比未精炼的检测器提高了 0.338 (p = 0.0015)，且仅异常增强目标一项在消融实验中就贡献了 0.463。我们还刻画了该方法适用的场景：增益取决于可用的连续且相互冗余的通道以及中断所移除的异常证据；而在以独热命令通道为主的遥测数据集和 PSM 数据集上，这种增益则不存在。

At ten sampling steps TADR runs in 41.4 ms per window on a single CPU core, sustaining about 24 concurrent 1 Hz streams from a 4.03 MB model. Code and per-seed results are available at this https URL.

在十个采样步骤下，TADR 在单核 CPU 上每窗口运行耗时 41.4 毫秒，能够通过一个 4.03 MB 的模型维持约 24 个并发的 1 Hz 数据流。代码和各随机种子的结果可在该 URL 获取。