Diffuse to Detect: A Generalizable Framework for Anomaly Detection with Diffusion Models Applications to UAVs and Beyond

TL;DR

The paper introduces Diffuse to Detect (DTD), a rapid and scalable anomaly detection framework using diffusion models, graph neural networks, and a two-branch architecture, demonstrating superior performance across various data types.

Contribution

It presents a novel single-step diffusion-based anomaly detection method with a flexible architecture, integrating GNNs and statistical techniques for enhanced accuracy and interpretability.

Findings

Outperforms existing anomaly detection methods on UAV sensor data

Effective across diverse data modalities including time series and images

Provides a scalable and interpretable framework for safety-critical applications

Abstract

Anomaly detection in complex, high-dimensional data, such as UAV sensor readings, is essential for operational safety but challenging for existing methods due to their limited sensitivity, scalability, and inability to capture intricate dependencies. We propose the Diffuse to Detect (DTD) framework, a novel approach that innovatively adapts diffusion models for anomaly detection, diverging from their conventional use in generative tasks with high inference time. By comparison, DTD employs a single-step diffusion process to predict noise patterns, enabling rapid and precise identification of anomalies without reconstruction errors. This approach is grounded in robust theoretical foundations that link noise prediction to the data distribution's score function, ensuring reliable deviation detection. By integrating Graph Neural Networks to model sensor relationships as dynamic graphs, DTD effectively captures spatial (inter-sensor) and temporal anomalies. Its two-branch architecture, with parametric neural network-based energy scoring for scalability and nonparametric statistical methods for interpretability, provides flexible trade-offs between computational efficiency and transparency. Extensive evaluations on UAV sensor data, multivariate time series, and images demonstrate DTD's superior performance over existing methods, underscoring its generality across diverse data modalities. This versatility, combined with its adaptability, positions DTD as a transformative solution for safety-critical applications, including industrial monitoring and beyond.