ChronosAD: Leveraging Time Series Foundation Models for Accurate Anomaly Detection

2026-05-31

AI Summary

This work addresses the limited generalization of time series anomaly detection in cross-domain scenarios and its difficulty in identifying subtle or context-dependent anomalies. To this end, the authors propose a two-stage architecture: first, a time series foundation model extracts generic embeddings in a zero-shot manner; then, a novel temporal module—integrating BiLSTM with multi-head attention—refines these embeddings to effectively capture long-range dependencies and enhance anomaly representations. Notably, the method achieves strong cross-domain generalization across industrial, healthcare, cyber-physical, and automotive systems without requiring task-specific fine-tuning. Evaluated on 11 benchmark datasets, it yields an average improvement of 4.72% in AUC and 6.60% in average precision, significantly outperforming existing approaches.

Abstract

Time series anomaly detection is a crucial task in various domains, including finance, healthcare, and industry. However, existing methods often struggle to generalize across different datasets, especially when anomalies are subtle or context-dependent. To solve this issue, we introduce ChronosAD, a novel architecture for anomaly detection that uses a time series foundation model as a feature extractor. Specifically, it employs a two-stage pipeline: first, it uses the foundation model to extract embeddings for each time series in a zero-shot manner. Then, a custom-developed Temporal Block, composed of Bidirectional Long Short-Term Memory (BiLSTM) and Multi-Head Attention, refines these embeddings to capture temporal dependencies and highlight salient patterns. Unlike previous approaches, our model requires minimal task-specific tuning and demonstrates robust generalization across a wide range of domains, including industrial, medical, cyber-physical, and automotive systems. Extensive experiments on 11 benchmarks show that ChronosAD outperforms existing methods by 4.72% in AUC and 6.60% in AP on average. The source code is available at https://github.com/intelligolabs/ChronosAD.

Problem

• time series anomaly detection
• generalization
• subtle anomalies
• context-dependent anomalies

Innovation

• time series foundation model
• anomaly detection
• zero-shot embedding
• Temporal Block
• BiLSTM with Multi-Head Attention