Dual Conditioned Motion Diffusion for Pose-Based Video Anomaly Detection

TL;DR

This paper introduces Dual Conditioned Motion Diffusion, a novel pose-based video anomaly detection framework that combines reconstruction and prediction advantages, utilizing conditioned motion and embedding, a motion transformer, and a new regularization to improve detection accuracy.

Contribution

The paper proposes a new framework called DCMD that integrates conditioned motion and embedding, a motion transformer, and a regularization technique for improved pose-based video anomaly detection.

Findings

Outperforms state-of-the-art methods on four datasets.

Demonstrates superior generalization in anomaly detection.

Effectively utilizes conditioned motion and embedding for better accuracy.

Abstract

Video Anomaly Detection (VAD) is essential for computer vision research. Existing VAD methods utilize either reconstruction-based or prediction-based frameworks. The former excels at detecting irregular patterns or structures, whereas the latter is capable of spotting abnormal deviations or trends. We address pose-based video anomaly detection and introduce a novel framework called Dual Conditioned Motion Diffusion (DCMD), which enjoys the advantages of both approaches. The DCMD integrates conditioned motion and conditioned embedding to comprehensively utilize the pose characteristics and latent semantics of observed movements, respectively. In the reverse diffusion process, a motion transformer is proposed to capture potential correlations from multi-layered characteristics within the spectrum space of human motion. To enhance the discriminability between normal and abnormal instances, we design a novel United Association Discrepancy (UAD) regularization that primarily relies on a Gaussian kernel-based time association and a self-attention-based global association. Finally, a mask completion strategy is introduced during the inference stage of the reverse diffusion process to enhance the utilization of conditioned motion for the prediction branch of anomaly detection. Extensive experiments on four datasets demonstrate that our method dramatically outperforms state-of-the-art methods and exhibits superior generalization performance.