Anomaly detection has been an active research area with a wide range of
potential applications. Key challenges for anomaly detection in the AI era with
big data include lack of prior knowledge of potential anomaly types, highly
complex and noisy background in input data, scarce abnormal samples, and
imbalanced training dataset. In this work, we propose a meta-learning framework
for anomaly detection to deal with these issues. Within this framework, we
incorporate the idea of generative adversarial networks (GANs) with appropriate
choices of loss functions including structural similarity index measure (SSIM).
Experiments with limited labeled data for high-speed rail inspection
demonstrate that our meta-learning framework is sharp and robust in identifying
anomalies. Our framework has been deployed in five high-speed railways of China
since 2021: it has reduced more than 99.7% workload and saved 96.7% inspection
time.

这篇论文提出了一个基于元学习的异常检测框架，它集成了生成对抗网络 (GANs) 的思想，并利用适当的损失函数，包括结构相似度指数 (SSIM)，解决了 AI 时代中针对大数据异常检测的挑战，试验表明该框架在高速铁路检测方面表现尖锐和强健，已在中国五个高速铁路上使用，使工作量减少了 99.7%，检查时间节省了 96.7%。