Motion Semantics Guided Normalizing Flow for Privacy-Preserving Video Anomaly Detection

TL;DR

This paper introduces MSG-Flow, a hierarchical model for privacy-preserving video anomaly detection that captures semantic and detailed motion features using normalizing flows and vector quantization.

Contribution

It proposes a novel hierarchical framework combining vector quantized auto-encoders, Transformers, and normalizing flows for improved skeleton-based anomaly detection.

Findings

Achieves state-of-the-art AUC scores of 88.1% and 75.8% on benchmarks.

Effectively models hierarchical motion semantics for better anomaly discrimination.

Outperforms existing skeleton-based methods in privacy-preserving video anomaly detection.

Abstract

As embodied perception systems increasingly bridge digital and physical realms in interactive multimedia applications, the need for privacy-preserving approaches to understand human activities in physical environments has become paramount. Video anomaly detection is a critical task in such embodied multimedia systems for intelligent surveillance and forensic analysis. Skeleton-based approaches have emerged as a privacy-preserving alternative that processes physical world information through abstract human pose representations while discarding sensitive visual attributes such as identity and facial features. However, existing skeleton-based methods predominantly model continuous motion trajectories in a monolithic manner, failing to capture the hierarchical nature of human activities composed of discrete semantic primitives and fine-grained kinematic details, which leads to reduced discriminability when anomalies manifest at different abstraction levels. In this regard, we propose Motion Semantics Guided Normalizing Flow (MSG-Flow) that decomposes skeleton-based VAD into hierarchical motion semantics modeling. It employs vector quantized variational auto-encoder to discretize continuous motion into interpretable primitives, an autoregressive Transformer to model semantic-level temporal dependencies, and a conditional normalizing flow to capture detail-level pose variations. Extensive experiments on benchmarks (HR-ShanghaiTech & HR-UBnormal) demonstrate that MSG-Flow achieves state-of-the-art performance with 88.1% and 75.8% AUC respectively.