FirstDiff: One-Step Diffusion-Based Anomaly Detection for Multivariate Time Series via Initial Noise Prediction

2026-08-16

AI Summary

This study addresses the computational inefficiency and neglect of intermediate information in diffusion-based multivariate time series anomaly detection by proposing FirstDiff. Leveraging a Diffusion Transformer to model dependencies, this framework demonstrates for the first time that initial reverse noise contains sufficient anomaly-related information. By utilizing statistical distributions from a validation set, FirstDiff enables single-step inference, thereby overcoming the limitations of traditional multi-step reverse processes. Experimental evaluations across five benchmark datasets confirm that FirstDiff achieves state-of-the-art performance while reducing inference overhead from full reverse trajectories to a single network evaluation. Consequently, this approach significantly enhances detection efficiency without compromising accuracy, offering a practical solution to the scalability challenges inherent in current diffusion models for time series analysis.

Abstract

Diffusion models have recently shown strong potential for multivariate time-series anomaly detection by learning the distribution of normal data through iterative denoising. Existing diffusion-based approaches, however, typically perform anomaly detection after completing the reverse diffusion process, relying primarily on the final reconstructed signal and overlooking informative representations produced during denoising. This design incurs substantial computational cost and limits the use of intermediate diffusion information for anomaly detection. In this paper, we propose FirstDiff, a diffusion-based anomaly detection framework based on the observation that the predicted diffusion noise at the initial reverse-diffusion evaluation already contains sufficient information for accurate anomaly detection. FirstDiff models the statistical distribution of predicted diffusion noise under normal behavior using validation data, enabling anomaly inference from a single denoising-network evaluation rather than completing the reverse diffusion trajectory. To model complex temporal and inter-sensor dependencies, FirstDiff employs a Diffusion Transformer as the denoising backbone. Extensive experiments on five public benchmark datasets demonstrate that FirstDiff achieves state-of-the-art performance while reducing diffusion inference from the full reverse trajectory to a single denoising-network evaluation.

Problem

• Multivariate Time Series
• Anomaly Detection
• Diffusion Models
• Computational Cost

Innovation

• One-Step Diffusion
• Initial Noise Prediction
• Anomaly Detection
• Multivariate Time Series
• Diffusion Transformer