用多尺度编码和动态代码本提升时序异常检测精度
COMET: Codebook-based Online-adaptive Multi-scale Embedding for Time-series Anomaly Detection
- 多尺度分块编码捕捉时间依赖与变量关联
- 双评分机制结合量化误差与记忆距离提升检出率
- 推理时在线更新代码本，适应分布变化
时序异常检测在工业领域至关重要。然而，现有方法对分块级表示中时间依赖性和多变量相关性的建模仍不充分，且依赖单一尺度模式，难以检测跨不同时间范围的异常。此外，仅关注正常数据表征使模型在推理时易受分布偏移影响。为此，我们提出面向时序异常检测的基于代码本的在线自适应多尺度嵌入方法（COMET），包含三个核心组件：(1) 多尺度分块编码，捕获跨多个分块尺度的时间依赖性与变量间相关性；(2) 向量量化核心集，通过代码本学习代表性正常模式，并利用量化误差与记忆距离的双重得分检测异常；(3) 在线代码本自适应，基于代码本条目生成伪标签，通过对比学习在推理阶段动态调整模型。在五个基准数据集上的实验表明，COMET 在45项评估指标中取得了36项最佳性能，验证了其在多样化环境下的有效性。
原文摘要 · Abstract (English)
Time series anomaly detection is a critical task across various industrial domains. However, capturing temporal dependencies and multivariate correlations within patch-level representation learning remains underexplored, and reliance on single-scale patterns limits the detection of anomalies across different temporal ranges. Furthermore, focusing on normal data representations makes models vulnerable to distribution shifts at inference time. To address these limitations, we propose Codebook-based Online-adaptive Multi-scale Embedding for Time-series anomaly detection (COMET), which consists of three key components: (1) Multi-scale Patch Encoding captures temporal dependencies and inter-variable correlations across multiple patch scales. (2) Vector-Quantized Coreset learns representative normal patterns via codebook and detects anomalies with a dual-score combining quantization error and memory distance. (3) Online Codebook Adaptation generates pseudo-labels based on codebook entries and dynamically adapts the model at inference through contrastive learning. Experiments on five benchmark datasets demonstrate that COMET achieves the best performance in 36 out of 45 evaluation metrics, validating its effectiveness across diverse environments.
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。