Intention-aware Hierarchical Diffusion Model for Long-term Trajectory Anomaly Detection

2025-09-21

AI Summary

Existing trajectory anomaly detection methods struggle to simultaneously capture high-level intent and low-level behavioral details, resulting in inadequate modeling of normal trajectory diversity. To address this, we propose an intent-aware hierarchical diffusion model: a high-level inverse Q-learning module evaluates the plausibility of sub-goal intentions, while a low-level conditional diffusion model reconstructs sub-trajectories; these components are jointly optimized to accurately characterize complex normal trajectory distributions. This work pioneers the integration of inverse Q-learning with diffusion models, establishing an end-to-end hierarchical anomaly detection framework where anomalies are identified via sub-trajectory reconstruction error. Experiments on long-horizon trajectory datasets demonstrate that our method achieves up to a 30.2% improvement in F1-score over state-of-the-art approaches, significantly enhancing robustness and generalization capability in trajectory anomaly detection.

Technology Category

• Humans and AI: Human-Aware Planning and Behavior Prediction
• Computer Vision: Diffusion Models for Vision
• Search and Optimization: Learning to Search

Application Category

• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Long-term trajectory anomaly detection is a challenging problem due to the diversity and complex spatiotemporal dependencies in trajectory data. Existing trajectory anomaly detection methods fail to simultaneously consider both the high-level intentions of agents as well as the low-level details of the agent's navigation when analysing an agent's trajectories. This limits their ability to capture the full diversity of normal trajectories. In this paper, we propose an unsupervised trajectory anomaly detection method named Intention-aware Hierarchical Diffusion model (IHiD), which detects anomalies through both high-level intent evaluation and low-level sub-trajectory analysis. Our approach leverages Inverse Q Learning as the high-level model to assess whether a selected subgoal aligns with an agent's intention based on predicted Q-values. Meanwhile, a diffusion model serves as the low-level model to generate sub-trajectories conditioned on subgoal information, with anomaly detection based on reconstruction error. By integrating both models, IHiD effectively utilises subgoal transition knowledge and is designed to capture the diverse distribution of normal trajectories. Our experiments show that the proposed method IHiD achieves up to 30.2% improvement in anomaly detection performance in terms of F1 score over state-of-the-art baselines.

Problem

• Detecting anomalies in long-term trajectories with complex spatiotemporal dependencies
• Simultaneously modeling high-level intentions and low-level navigation details
• Capturing full diversity of normal trajectories for improved anomaly detection

Innovation

• Uses hierarchical diffusion model for trajectory analysis
• Leverages Inverse Q Learning for intention evaluation
• Combines subgoal transition knowledge with reconstruction error